Download Checkel, Jeffrey T. - Transnational Dynamics of Civil War-Cambridge University Press (2013)

|
Transnational Dynamics of Civil War
Civil wars are the dominant form of violence in the contemporary international system, yet they are anything but local affairs. This book explores
the border-crossing features of such wars by bringing together insights from
international relations theory, sociology, and transnational politics with a
rich comparative-quantitative literature. It highlights the causal mechanisms –
framing, resource mobilization, socialization, among others – that link the
international and transnational to the local, emphasizing the methods
required to measure them. Contributors examine specific mechanisms
leading to particular outcomes in civil conflicts ranging from Chechnya, to
Afghanistan, to Sudan, to Turkey. Transnational Dynamics of Civil War thus
provides a significant contribution to debates motivating the broader move to
mechanism-based forms of explanation, and will engage students and
researchers of international relations, comparative politics, and conflict
processes.
j e f f r e y t . c h e c k e l is Professor of International Studies and Simons
Chair in International Law and Human Security at Simon Fraser University,
and Research Professor in the Centre for the Study of Civil War at the Peace
Research Institute Oslo (PRIO). He has published extensively in leading
European and North American journals, and is the author of Ideas and
International Political Change: Soviet/Russian Behavior and the End of the
Cold War (1997), editor of International Institutions and Socialization in
Europe (Cambridge University Press, 2007), and co-editor (with Peter
J. Katzenstein) of European Identity (Cambridge University Press, 2009).
Transnational Dynamics
of Civil War
Edited by
jeffrey t. checkel
Simon Fraser University
Peace Research Institute Oslo
cambridge university press
Cambridge, New York, Melbourne, Madrid, Cape Town,
Singapore, São Paulo, Delhi, Mexico City
Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by
Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9781107025530
© Cambridge University Press 2013
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.
First published 2013
Printed and bound in the United Kingdom by the MPG Books Group
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data
Transnational dynamics of civil war / edited by Jeffrey T. Checkel, Simon Fraser
University, Peace Research Institute Oslo.
pages cm
Includes bibliographical references and index.
ISBN 978 1 107 02553 0
1. Civil War. 2. International relations. I. Checkel, Jeffrey T., 1959 editor
of compilation.
JC328.5.T73 2013
303.60 4 dc23
2012028899
ISBN 978 1 107 02553 0 Hardback
Cambridge University Press has no responsibility for the persistence or
accuracy of URLs for external or third party internet websites referred to
in this publication, and does not guarantee that any content on such
websites is, or will remain, accurate or appropriate.
For Bullers
Contents
List of figures
page ix
List of tables
x
List of contributors
xi
Preface
Part I
1
3
4
5
6
Civil war – mobilizing across borders
Transnational dynamics of civil war
Jeffrey T. Checkel
Part II
2
xiii
3
Transnationalized civil war
Copying and learning from outsiders? Assessing diffusion
from transnational insurgents in the Chechen wars
Kristin M. Bakke
31
Mechanisms of diaspora mobilization and the
transnationalization of civil war
Fiona B. Adamson
63
Refugee militancy in exile and upon return in Afghanistan
and Rwanda
Kristian Berg Harpviken and Sarah Kenyon Lischer
89
Rebels without a cause? Transnational diffusion and
the Lord’s Resistance Army, 1986–2011
Hans Peter Schmitz
120
Transnational advocacy networks, rebel groups, and
demobilization of child soldiers in Sudan
Stephan Hamberg
149
vii
viii
7
Contents
Conflict diffusion via social identities: entrepreneurship
and adaptation
Martin Austvoll Nome and Nils B. Weidmann
Part III
8
9
173
Theory, mechanisms, and the study of civil war
Causal mechanisms and typological theories in the study
of civil conflict
Andrew Bennett
205
Transnational dynamics of civil war: where do we
go from here?
Elisabeth Jean Wood
231
Bibliography
259
Index
295
Figures
Figure 7.1 Single model runs with two disconnected
countries, for low and high noise levels.
Figure 7.2 Diffusion by transnationalized social
adaptation, for transnational impact set to
0.4, under the low and high noise scenario.
Figure 7.3 Diffusion by transnationalized norm
entrepreneurship, under the low and high
noise scenario.
page 190
194
198
ix
Tables
Table 5.1 Transnational mechanisms and shifts in
framing, resources, and repertoires
Table 6.1 Alternative explanations and testable
implications
Table 7.1 Number of runs (out of 50) with successful
norm convergence in the baseline model
Table 7.2 Number of runs (out of 50) with successful
norm diffusion for different levels of
transnational impact
Table 7.3 Number of runs (out of 50) with successful
entrepreneur-driven norm diffusion
Table 8.1 A taxonomy of theories on social mechanisms
Table 8.2 Locating the causal mechanisms invoked in
the empirical chapters within the taxonomy
Table 8.3 A typological theory on triangular bargaining
among rebels, home states, and host states
x
page 124
157
191
196
197
215
218
226
Contributors
f i o n a b . a d a m s o n – Associate Professor, Department of Politics
and International Studies, SOAS, University of London
k r i s t i n m . b ak k e – Lecturer in Politics and International Relations,
School of Public Policy, University College London
a n dr e w b e nn e t t – Professor, Department of Government,
Georgetown University
j e f fr e y t . ch e c ke l – Professor and Simons Chair, School for
International Studies, Simon Fraser University, and Research Professor,
Centre for the Study of Civil War, Peace Research Institute Oslo
s t e p h a n h a m b e rg – Ph.D. Candidate, Department of Political
Science, University of Washington
k r i s t i a n b e r g h a r p v i k e n – Director, Peace Research Institute
Oslo
s a r a h k e ny o n l i s c h e r – Associate Professor, Department of
Politics and International Affairs, Wake Forest University
m a r t i n a u s t v ol l no m e – Post-doctoral Research Associate,
Centre for the Study of Civil War, Peace Research Institute Oslo
hans pe te r schmit z – Associate Professor, Maxwell School of
Citizenship and Public Affairs, Syracuse University
xi
xii
List of contributors
n i l s b . w e i d m a n n – Marie Curie Post-Doctoral Fellow, Centre for
the Study of Civil War, Peace Research Institute Oslo
e l i s a b e t h j e a n w o o d – Professor, Department of Political
Science, Yale University
Preface
Our iconic images of civil war emphasize its local and localized nature.
From village-level atrocities during the Greek civil conflict, to bombedout buildings in Grozny, the capital of Chechnya, to child soldiers
cradling AK-47s in Uganda, the focus is on failed or failing states.
The very term civil war is meant to mark such conflict as something
quite different from inter-state wars. Indeed, the word civil implies that
such wars play out in some reasonably well-defined space, typically
within the borders of a state.
We now know, however, this is rarely in fact the case. Rebel groups
fighting civil wars often operate and recruit across borders; in many
instances, neighboring states intervene in an ongoing civil conflict;
diaspora communities – transnational in nature by their very essence –
play key roles in many civil wars. Such wars thus have border-crossing
and transnational dimensions; these need to be explored – theoretically
and empirically – if we are to understand fully the complex dynamics
behind what has become the dominant mode of organized violence in
the international system.
This book addresses such challenges and does so by bringing
together insights and arguments from two distinct research programs:
work in international relations and sociology on peaceful transnationalism; and a rich and growing comparative-quantitative literature on
civil war. The latter has documented in a sophisticated and rigorous
manner the importance of transnationalism in civil conflict; the former
provides a set of theoretical arguments and methods to explain the
nature of these transnational dynamics. Put more colloquially, ours is
an endeavor where transnationalism meets the dark side of global
politics.
This meeting is captured through the articulation of several causal
mechanisms linking the international and transnational to the local
and particular. However, the analysis is anything but abstract, as we
examine specific mechanisms leading to particular outcomes in a given
xiii
xiv
Preface
civil conflict. This operational focus puts a premium on method, on
how we actually measure mechanisms in action. And here, we join
with many others in arguing that a key measurement tool is process
tracing. Thus, beyond literatures on transnationalism and civil war,
our approach and findings speak directly to key conceptual and methodological debates motivating the broader move to mechanism-based
forms of explanation.
All chapters have benefitted from several rounds of discussion and
revision. Chapter 1 started as a conceptual memo for a first project
workshop, a brainstorming meeting held at the Peace Research Institute Oslo (PRIO) in December 2007. This was followed by three
workshops – Washington DC (October 2008), Simon Fraser University
(October 2009), PRIO (August 2010) – where the other chapter drafts
were discussed and debated.
I owe thanks to many people and institutions, most importantly to
the Centre for the Study of Civil War (CSCW) at PRIO. Scott Gates,
Andrew Feltham, and others created a welcoming and intellectually
exciting environment for a civil-war “newbie” – one who had spent the
previous decade studying a slightly different topic (European integration and identity). I would like especially to thank and acknowledge
Kristian Berg Harpviken, both for inviting me to join the CSCW and
for his key role at early stages of this project.
Matthew Evangelista and Scott Gates gave indispensable help at a
crucial later stage. At our last workshop, they acted as discussants not
only of individual chapters, but also of the project as a whole. Their
differing areas of expertise (Evangelista – transnationalism; Gates – civil
war), trenchant criticisms, and constructive suggestions have made this
a much better book. Shortly after this final meeting, when the manuscript was under review at Cambridge, two anonymous reviewers did a
brilliant job reminding me that ours is a volume with important things to
say about civil war – and not just about method and mechanisms!
For helpful comments on various parts of the manuscript, I thank – in
addition to those already named – Andy Bennett, Lars-Erik Cederman,
Marty Finnemore, Stephan Hamberg, Anna Holzscheiter, Andy Mack,
Martin Austvoll Nome, Martha Snodgrass, and Elisabeth Wood, as
well as the students of my May 2011 Ph.D. seminar, “Qualitative
Methods and the Study of Civil War,” and seminar audiences at the
Peace Studies Program, Cornell University; the Mershon Center for
International Security Studies, Ohio State University; the Swiss Federal
Preface
xv
Institute of Technology (ETH Zürich); the University of Washington
International Security Colloquium; the Otto Suhr Institute of Political
Science, Free University Berlin; and the Liu Institute for Global Issues,
University of British Columbia.
Last and certainly not least, I owe a debt of gratitude to Damian
Penfold, who carefully – and cheerfully – edited and formatted the entire
manuscript in-between conducting stints, and to Eileen Doherty-Sil for
preparation of the index. At Cambridge University Press, I thank John
Haslam for organizing an efficient, rigorous but fair review process, and
Carrie Parkinson for overseeing the production of the book.
For administrative and logistical assistance, I thank Andrew Feltham
at the CSCW in Oslo, Ellen Yap at the School for International Studies
at Simon Fraser University, and the Moynihan Center at Syracuse
University. Financial support was provided by the Norwegian
Research Council, the Centre for the Study of Civil War at PRIO,
Syracuse University, and the Simons International Endowment at
Simon Fraser University.
I end by stating the obvious. This book would not exist if it were not
for the dedication and perseverance of its contributors. They not only
complied – many times – with my requests for changes and revisions,
they also helped a recovering “Euroholic” rediscover his roots in what
we used to call security studies. When I left the field 20 years ago, the
central actors were superpowers with their ICBMs; today, they are
more likely to be imploding states and rebel groups. It is my modest
hope this book tells us something new about the latter.
JTC
Vancouver
PART I
Civil war- mobilizing across
borders
|
1
Transnational dynamics of civil war*
jeffrey t. checkel
Civil war has become the dominant mode of organized violence in
the post-Cold War international system. Depending upon the counting
rule employed, such wars have afflicted from a third to a half of
all nations. This internal warfare is not just extremely common, it is
persistent, with 20 percent of nations experiencing at least ten years
of civil war since 1960. During the mid 1990s – to take just one
example – nearly a third of the countries in sub-Saharan Africa had
active civil wars or conflicts (Blattman and Miguel 2010, 3–4).
Equally important, this form of ‘internal warfare’ is rarely internal,
with 55 percent of all rebel groups active since 1945 having transnational linkages (Salehyan 2009, 5). Indeed, civil wars nearly always
create opportunities and incentives for outside actors to intervene;
these actors may be other states, rebel groups, transnational civil
society, or the international community, and this intervention may
be malign (fanning the war) or benign (transnational NGOs targeting
the use of child soldiers). Moreover, such wars are often fueled by
cross-border flows of goods, including material (weapons), money
(diaspora financing) and human (new recruits for rebel groups).
Finally, civil wars can spur social mobilization across borders – by
strengthening senses of community among ethnic co-brethren, say.
So, the transnational clearly matters, with scholars documenting a
strong correlation between various transnational factors and actors
and changes in civil war dynamics. In this important sense, such wars
are no different. Across a variety of subfields and research programs in
comparative politics and international relations (IR), it has become a
* For helpful comments and discussions, I thank the project participants, the
students of my May 2011 Ph.D. seminar, “Qualitative Methods and the Study of
Civil War,” and, especially, Andy Bennett, Lars Erik Cederman, Matt
Evangelista, Marty Finnemore, Scott Gates, Stephan Hamberg, Andy Mack,
Martin Austvoll Nome, Martha Snodgrass, Elisabeth Wood, and two
anonymous reviewers for Cambridge University Press.
3
4
Jeffrey T. Checkel
truism to argue that the external and the internal, the global and the
local, the state and non-state actors are inextricably linked. The theoretical challenge – for scholars in general and students of civil war in
particular – is to explain the interactions across these various levels.
This introduction and the essays that follow take up this challenge,
exploring the relation of the transnational to the local in the context of
civil war. How do we conceptualize this transnational dimension? In
material or social terms? How does it affect civil war dynamics? By
bringing new material resources into play? By affecting cost/benefit
calculations? By promoting learning among actors? Under what conditions do transnational factors increase or decrease levels of civil violence?
What is the nature of the causal connection between the transnational
and the local? Put differently, what is the causal mechanism at work?
We argue that to address these issues requires three moves. Theoretically, the finding of transnationalism’s importance in civil war needs to
be linked to existing literatures in other subfields that have extensively
conceptualized and empirically documented such non-state dynamics;
key here is work on transnational politics in IR theory and sociology.
Analytically, one needs a more robust understanding of causality,
where the goal is the measurement of causal mechanisms and not simply
establishing causal effects. Methodologically, the central challenge is
practical – to measure mechanisms in action.1
The volume thus addresses gaps and promotes learning across three
literatures. For students of civil war, we supplement political economy
models and correlational analysis with process-based evidence on its
social and transnational dimensions. We thus provide new insights to
enduring questions related to civil conflict – agency and motives, group
mobilization, and international intervention, to name just a few. For
those studying transnationalism, we build upon but go beyond a focus
on the benevolent side of world politics by exploring and theorizing
transnational violence – that is, cross-border activities with malevolent
intent and consequences. For scholars interested in process, we provide
detailed evidence for the advantages – and disadvantages – of a move
to mechanism-based theorizing.
The remainder of this introductory essay is organized as follows.
I begin with a brief review of work on transnationalized civil war,
arguing that a diffusion metaphor is often invoked to link the external
1
More formally, we measure their observable implications; see below.
Transnational dynamics of civil war
5
and internal. To operationalize it, I draw upon – in the second section –
theories of transnational politics. Work by international relations
specialists and sociologists offers a rich menu of mechanisms –
framing, learning, brokerage, persuasion – to explain how transnationalism matters in civil war. The third section connects the theory to
data by focusing on method; special attention is paid to process
tracing, as this is particularly well suited to measuring causal mechanisms. In the final section, I conclude and preview the volume’s
structure.
Civil war and the transnational
Over the past decade, new research on civil war has put its study
squarely in the academic mainstream. At first quantitative in nature, it
has been complemented in recent years by a growing qualitative literature on civil conflict. My purpose here is not a wide-ranging review of
this rich literature – for this, see Tarrow 2007, and Blattman and
Miguel 2010; rather, the concern is how this work conceptualizes and
measures transnationalism (see also Wood, this volume).
Early efforts emphasized aggregate measures (Fearon and Laitin
2003), thus overlooking the sub-national, international, or transnational
dimensions of civil war. There was an inclination “to treat civil wars as
purely domestic phenomena” and a consequent neglect of “transborder
linkages and processes” (Cederman, Girardin, and Gleditsch 2009,
404). More generally, the analytic starting point was individual states
treated as independent entities – a so-called closed polity approach
(Gleditsch 2007). Cognizant of this limitation, several scholars spearheaded a two-pronged move to develop more disaggregated databases.
Some have disaggregated geographically and spatially, using so-called
geo-referenced conflict data (Buhaug and Gates 2002; Buhaug and
Rød 2006).
More important for my purposes, others have disaggregated by
moving away from state-level, aggregate proxies – for example,
by coding the attributes of non-state conflict actors (Cunningham,
Gleditsch, and Salehyan 2006). This has allowed researchers to
document the impact of new actors and interactions across state
boundaries in a wide array of cases. Work of this sort is important,
not only advancing the civil-war research program, but also – by
adopting an open polity perspective – aligning itself with the bulk of
6
Jeffrey T. Checkel
IR scholarship. It has allowed scholars to offer a more nuanced
picture of civil conflict, including its transnational dimensions.
As one important example, consider Salehyan’s book-length study
on what he calls transnational insurgencies (Salehyan 2009). In it, he
utilizes rich and disaggregated data embedded in a rigorous research
design to document that external sanctuaries have played a central
role in more than half of armed insurgencies since 1945 (see also
Malet 2010). Salehyan also explains exactly how international borders
shape the behavior of rebel groups involved in such conflicts. Simply
put, insurgencies seeking to challenge a state often have the option of
mobilizing abroad – and especially in neighboring states – where they
are safely out of reach of domestic foes. The book is anything but a
closed polity approach; rather, it demonstrates the analytic power to
be gained when civil wars are viewed in their broader international
and transnational contexts (see also Kalyvas and Balcells 2010).
Theory and analytics
Scholarship on the transnational dimensions of civil conflict has illuminated the broad trends at work, and has done so in a methodologically
self-conscious way that enhances the reliability of the findings. This is
no small feat, especially when one remembers that the data come from
violent and inaccessible parts of the world, from failed or failing states.
Moreover, there has been a concern to integrate the data with theoretical arguments of increasing sophistication – for example, Salehyan’s
theory of transnational rebellion that draws upon work on political
opportunity structures and conflict bargaining (Salehyan 2009, ch. 1).
This is a welcome and progressive move (Checkel 2010).
At the same time, such theorizing is typically not linked to the now
voluminous literature on transnationalism in world politics (Risse
2002). This seems odd – for a more systematic connection to this
research would alert scholars to alternative theoretical starting points
for understanding the cross-border dynamics of civil conflict (see also
Tarrow 2007, 588–590).
The work of Kristian Gleditsch and his collaborators is typical in
this regard. More than any others, they have documented the role
of various transnational factors in civil war (Gleditsch and Salehyan
2006; Gleditsch 2007). It is somewhat puzzling, however, that these
effects are interpreted primarily through the lens of rational choice,
Transnational dynamics of civil war
7
where transnationalism influences the dynamics of civil war by altering
material incentives and shifting cost/benefit calculations (Gleditsch
and Salehyan 2006, 341; Gleditsch 2007, 294–299; see also Gates
2002; Fortna 2004, 487–490; Kalyvas 2006, 2008; Humphreys and
Weinstein 2007; Blattman 2007, 7–13; Toft 2007, 102–107). Part of
the justification for this focus is that non-material factors such as norms,
ideas, and learning are difficult to measure (Salehyan 2009, 39)2 – an
assertion that is hard to understand in light of advances in empirical
constructivist research over the past 15 years (Adler 2002).
The literature on transnational politics is much more agnostic
on social theory, with some preferring rational choice (Cooley and
Ron 2002), some social constructivism (Price 1998), while still others
combine the two (Risse, Ropp, and Sikkink 1999a). If the goal is to
understand the full range of dynamics affecting civil war, then we
can only gain by adopting a broadened social-theoretic starting point.
As Cederman, et al., argue “additional research is needed on the details
of the border-transgressing bond, especially as regards the nature of
the actor-specific mechanism” (Cederman, Girardin, and Gleditsch
2009, 433). And from a problem-driven perspective, this ‘bond’ may
equally well be captured by rationalist or constructivist perspectives.
Methodological issues
Students of civil war have shown a growing interest in process (Tarrow
2007) – for example, in exploring the linkages from the transnational
to a particular civil conflict. Given this shift, it is not surprising that
the language of causal mechanisms is now often invoked (Fortna 2004;
Gleditsch and Salehyan 2006, 335–336, 360; Salehyan 2008). Yet, it
is unclear what is meant by such language. In one recent study where
the transnational to civil war nexus is explored, causal mechanisms
seem central to the analysis, but are never defined. Instead, it is left to
the reader to infer that a mechanism equals a hypothesis, diffusion,
spillover effects, or ethnicity (Cederman, Girardin, Gleditsch 2009,
408, 412, 433; see also Gleditsch 2007, 297; and the discussion in
Fjelde “Transnational Dimensions,” 5). Causal mechanisms can be
2
See also Hanne Fjelde. (undated) “Transnational Dimensions of African Civil
Wars and the Triple R Framework.” In Thomas Ohlson (ed.), From Intra State
War to Durable Peace: Conflict and its Resolution in Africa after the Cold War.
Unpublished Manuscript, Uppsala University, 14 15.
8
Jeffrey T. Checkel
defined as the process by which an effect is produced or a purpose is
accomplished (see next section). However, it is not clear this is what
the civil war scholars are measuring empirically. Moreover, even in
those cases where more effort is devoted to theorizing transnational
mechanisms, these then vanish in the empirical testing (Gleditsch and
Salehyan 2006, 342, 347–350). There would appear to be a mismatch
between the language of causal mechanisms and methodological
choice, between conceptualizing cause as a process (! mechanisms)
and measuring it via quantitative techniques (! covariation).
Even the most sophisticated, mixed method research on transnationalized civil war has problems at this level. Consider again Salehyan’s
Rebels without Borders (2009). Methodologically, the rigor of the
book’s quantitative first half is not carried over to its case studies, which
is unfortunate given the availability of an increasingly sophisticated
case-methods literature (Bennett and George 2005; Gerring 2007a, for
example). As a result, the case studies – despite the author’s claims
(Salehyan 2009, 108, 110, 122–123, 164) – fail to provide evidence
for the causal mechanisms he posits to explain the correlational findings.
For example, bargaining dynamics based on cost/benefit calculations
figure prominently in Salehyan’s theory; yet, the case study chapters
provide no process-level evidence that such dynamics were at work.
Here again, the transnational politics literature offers both conceptual discussions and empirical applications relevant to students of civil
war. Conceptually, it has taken seriously the logical implications of
open polity models, developing cross-level theoretical frameworks.
These move well beyond level-of-analysis approaches (Singer 1961)
or arguments about residual variance (Moravcsik 1993), to emphasize
cross-level interactions that put the spotlight on process (Keck and
Sikkink 1998; Risse, Ropp, and Sikkink 1999a). Such a conceptual
move is necessary if one is to explore the role of transnational causal
mechanisms.
Empirically, scholars of transnationalism have put a good bit of
thought into the methods one needs to measure such dynamics. Moving
beyond the measurement of causal effects, they have demonstrated that
techniques are available for capturing mechanisms. In particular, they
have convincingly shown the utility of a method known as process
tracing (Risse-Kappen 1995a; Risse, Ropp, and Sikkink 1999a), which
is seen as key for measuring mechanisms (Bennett and George 2005,
ch. 10). These empirical, operational applications of the technique
Transnational dynamics of civil war
9
should be of great help to students of civil war, who increasingly
invoke it while failing to demonstrate how it works in practice
(Weinstein 2007, 53–59; see also Kalyvas 2007).
Summary
Scholarly research has documented that transnationalism plays an
important role in civil conflict, but the specific causal mechanisms
remain poorly understood, for both theoretical and methodological
reasons. From a broader perspective, these limitations are understandable and no surprise. The success to date of the civil war research
program (Blattman and Miguel 2010) means these studies have
reached a new level of sophistication, where open-polity models and
transnational dynamics are stressed, and causal mechanisms are
invoked. However, it is one thing to speculate about transnational
mechanisms; it is quite another properly to theorize and measure them.
Transnational mechanisms of civil war
Progress in addressing these gaps requires a two-fold analytic-theoretical
move – to the language and practice of causal mechanisms and to
theories of transnationalism. I begin by situating this volume’s
understanding of causal mechanisms in the (vast) literature on them.
Next, I turn to transnationalist scholarship in international relations
and sociology. This work suggests specific ways to make operational
the diffusion mechanism invoked in much of the civil war literature,
a point I demonstrate with examples from the volume’s empirical
contributions. A third section then turns the tables, suggesting that
transnationalists have something to learn as well, in particular, how
to theorize transnational violence.
Causal mechanisms – from confusion to emergent consensus
Thinking about mechanisms has a long history in the philosophy of
science and in the social sciences. Philosophers – for several hundred
years – and social scientists – more recently – have debated the nature
and meaning of cause (Kurki 2008 for a state-of-the-art review). Is it
best captured by a Humean understanding of constant conjunction
and covariation or a realist account of cause as process? Among
sociologists and inspired by the work of Robert Merton and his
10
Jeffrey T. Checkel
colleagues at Columbia University, mechanisms were the subject of
intensive inquiry in the early years after World War II (Hedström and
Swedberg 1998, 1–2). Thus, in the distant and not-so-distant past, the
interest in mechanisms was there.
Such interest has blossomed among political scientists over the past
decade, due both to a growing dissatisfaction with structural theories
(Hall 2003, 375–388) and the rise of a new generation of scholars (see
also Bennett, this volume). Rationalists now do mechanisms (Elster
1998), constructivists see them as a core component of their social
theory (Wendt 1999), quantitative researchers increasingly invoke
them (Gleditsch and Salehyan 2006, 341–344), and – among qualitative
methodologists – new research on case studies gives mechanisms a pride
of place (Bennett and George 2005; Gerring 2007a).
One not surprising result of all this attention is that different authors
define a causal mechanism in different ways, a fact now widely noted
and bemoaned (Mahoney 2001; Gerring 2007b; Falleti and Lynch
2009). At an intuitive level, it is easy to define a mechanism – it
connects things and captures process. However, the devil, as always,
is in the detail. Are mechanisms easy or hard (i.e., unobservable) to
see? Must the use of mechanisms be premised on an ontological stance
of methodological individualism?
Building upon recent discussions in the literature, I define a causal
mechanism as “the pathway or process by which an effect is produced
or a purpose is accomplished” (Gerring 2007b, 178; see also Caporaso
2009). Mechanisms are thus “relational and processual concepts . . .
not reducible to an intervening variable” (Falleti and Lynch 2009,
1149); they are “the operative or motive part, process or factor in a
concrete system that produces a result” (Wight 2006, 34). These
minimalist definitions capture other extant usages of the term (Gerring
2007b; see also Gerring and Barresi 2003).
Moving from conceptualization to a more operational level, there is
also a growing consensus on measurement. Philosophers of social
science – and especially scientific realists – view causal mechanisms
as ultimately unobservable ontological entities that exist in the world,
not in our heads; they are thus more than mere “analytical constructs
that facilitate prediction” (Wight 2006, 31–32, quote at 31). If
mechanisms are real but unobservable entities, the implications for
measurement are clear: we measure not hypothesized mechanisms,
but their observable implications.
Transnational dynamics of civil war
11
Given the invocation of scientific realism in the last paragraph,
I should highlight one area where considerable confusion remains:
the philosophical foundation of empirical, mechanism-based social
science. As a number of sharp-thinking analysts have noted, there is
a tension between the growing use of mechanism-based thinking in
American political science and this same community’s continuing
adherence to a philosophical position – positivism – at odds with a turn
to mechanisms (McKeown 1999, 163–164; Johnson 2006; Wight 2006,
31–32; Gerring 2007b, 164). For sure, such philosophical disputes,
by their very nature, will not be resolved any time soon (if ever).
In this volume, our response to such confusion and disputes is twofold. First and pragmatically, we just get on with it, doing methodologically self-conscious mechanism-based social science. Implicitly,
all contributors adopt a variant of scientific realism. With its emphasis
on measuring causation via mechanisms and a practical stance of
“epistemological opportunism” (Wight 2002, 36), it well suits
our purposes. The latter point is important, as methods flow from
epistemology. This opportunism thus legitimates and indeed mandates
a plurality of methods when it comes to the measurement of mechanisms. Contributors to this volume thus measure mechanisms in a
variety of ways, from case studies and process tracing to agent-based
modeling.
Second and conceptually, we step back and use our results to assess
the fit between research practices and their philosophical foundations
(Bennett, this volume). Is scientific realism really such a good basis for
what we claim to be doing, especially given the strong critiques recently
leveled against it (Chernoff 2002)? Would a form of pragmatism or
analytic eclecticism be better (Sil and Katzenstein 2010a, 2010b)?
Whatever the philosophical foundation, can we articulate clear and
agreed community standards for what counts as good mechanismbased social science?
Transnational politics and civil war
Students of transnational politics – mainly in political science and
sociology – have been thinking about international-domestic connections for quite some time. Over nearly four decades, this work has gone
through three distinct phases. The earliest research simply challenged
realist assumptions about states as the key actors in international
12
Jeffrey T. Checkel
politics, getting transnationalism on the scholarly agenda (Keohane
and Nye 1972). A second generation, appearing in the 1990s, marked
a significant analytical shift – exploring when and under what conditions transnationalism mattered (Risse-Kappen 1995a; Evangelista
1999; Risse, Ropp, and Sikkink 1999a).
More recently, a third set of scholars has disaggregated key assumptions about transnational actors and domestic politics. They have begun
to look inside entities such as NGOs, asking what motivates them to
act (Schmitz 2004) and why they mobilize around some issues but not
others (Carpenter 2007). If earlier research stressed ideational and normative motivations, this new work sees transnational actors as highly
strategic and calculating as well (Cooley and Ron 2002; della Porta and
Tarrow 2004; Bob 2005). On domestic politics, newer work theorizes
it in greater detail and does so more systematically, thus avoiding the
ad-hocism that often prevailed in earlier research (Orenstein and
Schmitz 2006; Schmitz 2006).
These multiple disaggregation moves have led to a greater interest
in the different kinds of causal mechanisms connecting transnational
actors and factors and domestic change (Price 1998; Tarrow 2001;
Tarrow and della Porta 2004; James and Sharma 2006; Checkel 2007;
see also Symposium 2006). Indeed, we now have a growing roster
of transnational causal mechanisms, including emulation, persuasion,
flows of resources (ideational or material), framing, and power
transitions – to name just a few.
Students of civil war can benefit in four specific ways from this
IR/sociology work on transnational causal mechanisms. First, it
highlights that mechanisms can have varying social-theoretic bases.
Some are captured by rational choice – power transitions; others by
constructivism – persuasion; and yet still others by both social
theories – learning in its simple and complex variants. This socialtheoretic pluralism would align civil-war studies with the philosophy
of science literature, where it is argued that accounts employing
causal mechanisms are in principle “quite compatible with different
social theories of action” (Mayntz 2004, 248; see also Mahoney
2001, 581).
Second, this roster of causal mechanisms offers opportunities for
better specification of the diffusion metaphor invoked in the civil-war
literature. To begin, what exactly is transnational diffusion? Research
on the topic defines it as the spread of policies, institutions, or beliefs
Transnational dynamics of civil war
13
across borders (Simmons, Dobbin, and Garrett 2006, 787–790).3 On its
own, however, the term obscures more than it reveals, a point now well
appreciated in both the transnationalist and policy diffusion literatures
(Symposium 2006). Instead, the cutting edge is to unpack diffusion and
explore the underlying causal mechanisms – coercion? learning?
emulation? – driving it (see also Tarrow 2010). Indeed, it is precisely
the latter that are hinted at but not theorized or tested in recent
work exploring the transnational dimensions of civil war (Cederman,
Girardin, Gleditsch 2009, passim).
Third, the mechanisms cited above have undergone extensive empirical testing (Risse-Kappen 1995a; Keck and Sikkink 1998; Price 1998;
Risse, Ropp, and Sikkink 1999a; Risse 2000; Lynch 2002; Checkel
2003, 2007; della Porta and Tarrow 2004; Schmitz 2006; see also
Levy, Young, and Zuern 1995; Johnston 2001, 2008; Gheciu 2005;
Lewis 2005). For example, scholars have now applied a number of
techniques for measuring these mechanisms – panel interviews,
surveys, text analysis, process tracing – while maintaining a healthy
appreciation of their limits (Gheciu 2005; Johnston 2008). And,
again, this research is neutral in social-theoretic terms. Some use
these methods to document rationalist cost/benefit causal mechanisms
(Schimmelfennig 2005), while others use them to measure noninstrumental constructivist learning (Checkel 2001).
Theorizing non-instrumental mechanisms may be especially helpful
as scholars of civil war continue their explorations of transnationalism’s
effects. To date, they have neglected the role of “noninstrumental
factors, such as norms and emotions” (Kalyvas 2006, 13; see also
Blattman and Miguel 2010, 18), or – specifically on rebel groups – have
been “silent on the internal psychology of the recruit,” thus making it
difficult to uncover “the root causes of . . . indoctrination” (Blattman
2007, 25–26, 27; see also Annan, et al. 2009; and Wood 2008, 2009).
Those root causes – to take just one example – may be partly captured
by the non-instrumental socialization mechanisms (role playing,
persuasion) emphasized by the transnationalists (Checkel 2007).
Fourth, the work by transnationalists provides students of civil
war the raw material for developing better mechanism-based theories
that are bounded, contingent, but still ‘small g’ generalizable (George
3
Wood, this volume, provides an excellent analysis of the broader diffusion
literature and its utility to students of civil war.
14
Jeffrey T. Checkel
1993; Hall 2003; Gerring 2007b). After some false starts – in particular,
vague invocations of middle-range frameworks – scholars are thinking
hard about how to produce rigorous mechanism-based theory (Elster
1998; Gates 2008; Lichbach 2008) and exploring its requirements in
terms of research design (Johnston 2005; Wood 2008, 556). Recently –
for example - they have proposed so-called typological theory,
where combinations of mechanisms interact in shaping outcomes
for specified populations (Bennett, this volume; see also Bennett and
George 2005, ch. 11).
If the foregoing suggests an “in principle” contribution from the
IR/sociology transnationalists, several examples from the chapters
that follow indicate its practical utility. In her study of the Chechen
civil wars, Kristin Bakke theorizes and traces the diffusion mechanisms
through which transnational insurgents affect domestic challengers to
the state. Via mediated or relational diffusion, these insurgents engendered learning and emulation among key Chechen rebel groups, in
turn (re-)shaping domestic processes of mobilization (Bakke, this
volume; see also Adamson, this volume, for a related argument as
applied to diasporas). Hans Peter Schmitz takes a hard case for external
influences – the Lord’s Resistance Army (LRA) in the Ugandan civil
conflict – and demonstrates that even here transnational emulation and
social adaptation played roles in shaping its behavior and evolution
(Schmitz, this volume).4
Turning to a very different type of transnational influence, Stephan
Hamberg asks whether transnational advocacy networks have had any
real influence on rebel groups’ use of child soldiers. Examining the case
of the Sudan People’s Liberation Army (SPLA), who in the early 2000s
demobilized several thousand child soldiers, he argues that social
shaming – the most common mechanism of transnational activism –
failed to induce a change in rebel group behavior. Rather, child soldiers
were demobilized – including by the SPLA – only when the international community granted material concessions or at least promised
them to rebel groups (Hamberg, this volume).
A final example returns to the diffusion of civil conflict, but employs
computer techniques instead of empirical case studies to capture the
causal mechanisms at work. In their chapter, Martin Nome and Nils
4
It is a hard case because the LRA’s long time leader, Joseph Kony, has always
sought to isolate it from external influences.
Transnational dynamics of civil war
15
Weidmann utilize agent-based modeling to analyze the diffusion of
social identities as a key process underlying the spread of civil conflicts.
They disaggregate – and thus better specify – diffusion as occurring
through two possible causal mechanisms: social adaptation in a transnational context, and transnational norm entrepreneurship. Their
simulations indicate that norm entrepreneurship is the more robust
mechanism of diffusion, which is an important confirmation of a
finding in the qualitative, IR/sociology literature.
These examples provoke two final – cautionary – remarks. First, we
well appreciate there is an issue of aggregation here. That is, while
cataloguing various transnational causal mechanisms is a useful
starting point, more is needed. After all, the broader international
relations literature is by now replete with such (non-cumulative) lists.
In fact, it is possible to construct an orderly taxonomy of theories on
the causal mechanisms of civil conflict – one that provides a comprehensive and useful checklist so scholars can ensure they are not leaving
out important mechanisms in their explanations. This is a task Andrew
Bennett takes up in his chapter (Bennett, this volume).
Second, we recognize that the IR/sociology work on causal mechanisms is not the definitive last word. One should maintain a healthy
skepticism for how well mechanisms theorized by scholars whose
prime interest is peaceful change travel to situations where violence
and institutional collapse are the norm. More positively, we can use
this translation exercise to push research on transnationalism in new
directions, as suggested below (see also Wood, this volume).
Theorizing transnationalized violence
The issue is how to theorize transnationalized violence. And here, the
IR/sociology transnationalists – to turn the tables – have much to learn
from work on conflict processes. They have focused on peaceful and
non-violent change in world politics (the spread of human-rights
norms, socialization by international institutions), thus failing to appreciate that their theories may mis-specify social dynamics in settings
marked by violent civil conflict.
Consider work on small group dynamics, which clearly play a role in
civil war – say, in the transborder recruitment of rebel group members.
Students of transnationalism have analyzed them through the prism of
socialization, or the process of inducting actors into the norms and
16
Jeffrey T. Checkel
rules of a given community (Risse, Ropp, and Sikkink 1999a). To do
this, they have drawn upon a particular strand of socialization literature, one where stable institutions are at work and power and coercion
are back-grounded (Checkel 2007).
Yet, this process of induction could surely work in more coercive
ways. For example, research on professional militaries and urban gangs
explores hazing, physical coercion, and even rape as causal mechanisms
leading to socialization. Such mechanisms might be highly relevant in
civil war settings, as Harpviken and Lischer argue in their chapter on
refugee return (Harpviken and Lischer, this volume; see also Wood
2009, 2010; Cohen 2010). The point is not to reject previous work by
transnationalists, but to expand their roster of causal mechanisms, thus
capturing the full array of social dynamics at work in the contemporary
transnationalized world – both its good and evil sides (Checkel 2011).
Summary
This volume brings together students of conflict processes, IR theorists,
comparativists, and methodologists to explore the evident fact that
civil wars are rarely contained within the borders of one country; they
have transnational dimensions that need to be captured. We ask two
questions. (1) What are the transnational mechanisms that influence
group mobilization during civil war? Under what conditions do they
enable organized violence? (2) What are the causal mechanisms at work
when transnational actors intervene to end violence in civil conflict?
These questions intentionally capture two key aspects of transnational influence on civil war. The first, a bottom-up perspective,
explores how actors in civil conflict may use or be affected by broader
transnational processes. The second is a top-down view, asking how
transnational actors may intervene to re-shape domestic conflict
dynamics. The difference between the two is subtle, but important
and essentially concerns the locus of agency. We do not a priori give
analytic priority to one or the other.
Providing operational, empirically grounded answers to these questions requires the following steps. First, we think in terms of transnational
causal mechanisms, which are defined – adapting Gerring (2007b) – as
the transnational pathway or process by which an effect is produced or
a purpose is accomplished. We theorize and test a number of causal
mechanisms relevant to transnationalized civil conflict.
Transnational dynamics of civil war
17
Second and regarding methods, we argue that there is no one technique that is best for measuring mechanisms. Contributors thus utilize
a broad range, from the purely qualitative, to cutting-edge methods
such as agent-based modeling. The mix varies, depending upon the
specific question asked and a contributor’s methodological proclivities.
Third, we critically evaluate our contributions. In part, this is done
in a standard way, as individual authors consider alternatives and
challenges to their arguments and evidence (see Hamberg, this volume,
for an excellent use of this strategy). However, we also step back and –
in two concluding essays – conduct net assessments. Conceptually, can
our findings be used to clarify the outlines of a cumulative civil-war
research program centered on mechanisms (Bennett, this volume)?
Theoretically, in what specific ways do we advance the research program
on civil war and its transnational dimensions (Wood, this volume)?
Causal mechanisms in action – the challenge
of measuring process
This section connects the analytics and theory to data by focusing on
method. The discussion proceeds in three stages, beginning with a brief
review of the best work exploring the methods-mechanisms nexus.
Next, I describe and assess the technique known as process tracing,
which seems particularly useful for measuring causal mechanisms.
Finally, I address feasibility. Can mechanism-based social science – with
its inherent requirement for significant data – be conducted in situations
where violence and institutional collapse are the norm?
State of the art
While progress is evident in methods work on mechanisms, there is
still a clear need to bring the discussion down to a more operational level –
how one actually uses mechanisms in empirical social science. Important
efforts by four sets of scholars – Bennett and George, Gerring,
McAdam-Tarrow-Tilly, and Wood – are indicative of the challenges
that lie ahead. If by methodology, one means how we come to know,
then the common denominator for these authors is that we come to know
about mechanisms by studying processes, dynamics, and narratives.
In their wide-ranging overview of the case-study approach, Bennett
and George devote an entire chapter to process tracing (Bennett and
18
Jeffrey T. Checkel
George 2005, ch. 10). This method “attempts to identify the intervening
causal process – the causal chain and causal mechanism – between an
independent variable (or variables) and the outcome of the dependent
variable” (Bennett and George 2005, 206). The chapter is exemplary in
noting process tracing’s compatibility with both rationalist and
constructivist thinking, in highlighting its different varieties, and in
giving several methodological tips to the novice.
However, this same novice will be frustrated at a practical level, as
basic questions sure to be on the lips of any first-time process-tracer
remain unanswered. How does one know when to stop with the
process tracing? When is there enough evidence to document the
workings of a particular causal mechanism – that is, what are the data
requirements (see also Gerring 2007a, 181)? What counts as good
process tracing? What are the community standards (see also Bennett
and Checkel forthcoming)?
More recently and from a methodologist’s perspective, Gerring
(2007a) has produced what critics are rightly calling the state of the
art on the case study approach. Like Bennett and George, he devotes an
entire chapter to process tracing. Yet, in an otherwise superbly crafted
book, this chapter is short and lacking in practical guidance on how to
execute the technique (see also Symposium 2007b, 5, 14). Indeed,
Gerring (and co-author Craig Thomas) despair of offering systematic
advice on process tracing, as it is as much “detective work” and
“journalism” as a rigorous social science method (Gerring 2007a, 178).
In light of the foregoing, the very title of McAdam, Tarrow, and
Tilly’s essay – “Methods for Measuring Mechanisms of Contention” –
would seem promising (McAdam, Tarrow, and Tilly 2008). Indeed,
they start by criticizing Bennett and George (2005) for failing “to tell
us how to describe – let alone measure – the causal mechanism or
mechanisms at the heart of the processes that interest them” (McAdam,
Tarrow, and Tilly 2008, 2–3).
In contrast to others, McAdam and co-authors get down to the brass
tacks of how to measure mechanisms, offering two direct measures
(events data, ethnography) and two indirect ones (comparative crossnational and intra-national data). This is excellent and very much
what work on causal mechanisms requires. However, by drawing
illustrations only from the contentious politics research program, the
utility of their analysis is limited. In addition, while the empirical
examples are interesting, it is difficult to extract a set of best
Transnational dynamics of civil war
19
methodological practices from them. Again, the novice – or aspiring
Ph.D. student – is left wondering how he/she actually does this
mechanism-based research (see also Symposium 2008).
Within the literature on civil war, Elisabeth Wood’s socialtheoretically plural and carefully documented work on the
Salvadoran civil war (Wood 2003) sets the standard for how to capture
the presence and role of causal mechanisms. Not only does she theorize
a wide range of mechanisms – ranging from (rationalist) calculations
of cost/benefit to (sociological-constructivist) morals, emotions, and
norms – Wood also provides rich evidence to document and measure
them. Her methods are also plural, ranging from ethnography to
formal modeling.
Yet this important work could be improved in several ways. At
an operational level, it is not entirely clear what procedures Wood
followed in reconstructing her group/individual mobilization dynamics.
How did she control for the bias generated by her largely micro,
individual perspective, where structural factors could be overlooked?
How did Wood decide at what level to focus her search for causal
mechanisms? How does one decide how close, how micro to go? This
problem of so-called infinite regress has been addressed in a general
sense (Gerring 2007b, 176), but with little practical advice on dealing
with it.
As Wood demonstrates empirically and as others argue from a more
methodological perspective (McAdam, Tarrow, and Tilley 2008;
Staggenborg 2008, 341), ethnography is important for measuring
process and mechanisms. Yet, its use raises two potentially vexing
issues. First, how did Wood know when to stop? Arbitrarily after 100
interviews? After 150? While the book provides no clear answer,
elsewhere she has argued – quite convincingly – that the ethnographic
field work continued until it revealed no new patterns. That is,
additional interviews were simply providing further evidence for the
operation of her core causal mechanisms (Wood, personal communication, June 2008; see also Gusterson 2008). More formally, one
could say that Wood was following a logic of Bayesian inference in
deciding when to stop (Bennett 2008).
Second, the epistemological foundation of ethnography –
interpretism – sits uneasily with the strong causal emphasis of
mechanism-based social science (see also Schatz 2009). How does one
square the circle here? Does it matter that Wood, within the covers of
20
Jeffrey T. Checkel
one book, utilizes methods – formalization, process tracing, ethnography – from very different epistemological traditions to measure her
mechanisms? She provides no clear answer – and is in good company.
In his commentary on McAdam, Tarrow, and Tilly’s (2008) analysis of
causal mechanisms, Lichbach (2008) hints at similar epistemological
tensions by invoking the oxymoronic phrase positivist constructivism
to describe their contribution. Clearly, this is an area in need of further
thought and reflection, one to which we return (Bennett, this volume).
In sum, recent years have seen growing attention to the operational
issue of how one measures causal mechanisms. This is a welcome and
healthy trend, moving scholars beyond conceptual discussions (what is
a mechanism, are they real, etc.). However, in a way quite similar to
the quantitative researchers in the civil-war literature, these students
of mechanisms are victims of their own success. Their expositions
and analysis are generally so clear, it is easy for (even sympathetic)
critics to see the remaining gaps and challenges.
Capturing mechanisms – process tracing
Recall our baseline definition of a causal mechanism: “the pathway or
process by which an effect is produced or a purpose is accomplished”
(Gerring 2007b, 178). Given this definition and the foregoing discussion, a number of methods seem relevant for documenting mechanisms.
These include ethnography, interview techniques, process tracing, and
even formal models; in addition, agent-based modeling is a promising
technique for exploring the logic and hypothesized scope conditions of
mechanisms (Nome and Weidmann, this volume; see also Cederman
2001; Weidmann 2007). Finally, statistical methods play a more indirect,
but still important role: establishing that there is a relation in the first
place that requires explanation.
Full detail on these techniques can be found in the chapters that
follow, as authors operationalize them in specific empirical contexts.
Here, I concentrate on practice, that is, how we have sought to craft
and implement mechanism-based empirical studies. After a few points
on design, I advance several evaluative criteria for accounts that
employ process tracing.
On design, a good mechanism-based explanation builds on four
components (Earl 2008, 356–357). As a pre-condition – step no. 1 –
one needs to be able to distinguish the elements from the mechanism(s)
Transnational dynamics of civil war
21
in the explanation, so we can study the mechanism. What is X and
Y and the mechanism that links them? Then – step no. 2 – we need
to show that relations among the elements were changed. More
important – step no. 3 – we need to show that the hypothesized
mechanism was responsible for altering that relationship and that
other, alternative mechanisms were not (see also the discussion of
equifinality below). So, for example, if diffusion (X) leads to conflict (Y),
we may hypothesize that learning was the mechanism explaining
that relation. Even more important, we then – step no. 4 – need to ask
in a very empirically grounded way “what are the observable implications
of learning?”
Bakke’s study of transnational diffusion and the Chechen civil wars is
an exemplary illustration of this design in practice. She operationalizes
such diffusion as – in part – mechanisms of learning and emulation;
develops operational indicators for them; and thinks temporally. For
the latter, it is crucial to ask whether domestic learning occurred after
intervention by transnational insurgents. Bakke also explores whether
alternative – domestic, in her case – mechanisms might have produced
the same outcome (Bakke, this volume).
Turning now specifically to process tracing, I argue it is possible to
articulate a three-part standard for what counts as a good application
of it (see also Symposium 2007b, 5; Bennett and Elman 2007, 183;
Bennett and Checkel forthcoming; Bennett, this volume). Metatheoretically, it will be grounded in a philosophical base that is
ontologically consistent with mechanism-based understandings of
social reality and methodologically plural, such as that provided by
scientific realism (Wight 2006, ch. 1), analytic eclecticism (Katzenstein
and Sil 2008; Sil and Katzenstein 2010b) or pragmatism (Johnson
2006). Contextually, it will utilize this pluralism both to reconstruct
carefully causal processes and not to lose sight of broader structuraldiscursive context. Methodologically, it will take equifinality seriously,
which means to consider the alternative causal pathways through
which the outcome of interest might have occurred.
With these three signposts in hand, good process tracing needs then to
address specific operational issues regarding theory, data, and method
(see also Checkel 1997, 2001, 2003, 2007, 2008a). Theoretically, the
study of causal mechanisms raises an infinite regress or stopping
point issue. When does the process tracing on mechanisms stop? How
micro should we go? In my own work on socialization, I took one
22
Jeffrey T. Checkel
mechanism – socialization – and broke it into three sub-mechanisms:
strategic calculation, role playing, and persuasion. These were then
subjected to process tracing (Checkel 2007). Why stop at this point,
though? Persuasion, for example, could be further broken down into its
own sub-mechanisms, most likely various types of cognitive processes.
Luckily, an answer to this “how micro” question is straightforward
and dictated by the state of disciplinary knowledge. In my case, this
indicated it was socialization – and not persuasion – that was ripe for
disaggregation into smaller component mechanisms. In addition and at
a more practical level, how micro one goes is also a function of
available data. It makes little sense to spend time attempting to process
trace what cannot be measured empirically.
A second theoretical issue is causal complexity. Thinking in terms of
mechanisms abstracts from and simplifies the real world – less than is
often the case, but abstract it still does. Moreover, because of the high
data requirements, it is all too easy to conduct process tracing on only
one mechanism. Yet, in many cases, the outcome observed is the result
of multiple mechanisms interacting over time. There is no easy answer to
this dilemma; one cannot conduct process tracing on all possible candidate
mechanisms. However, thinking explicitly and early about equifinality,
where the outcome of interest may be the result of alternative causal
pathways (Bennett and George 2005, 161–162), and using agent-based
modeling, which allows for systematic exploration of possible interactions
among hypothesized mechanisms (Nome and Weidmann, this volume;
Cederman 2003, 146; Hoffmann 2008), can bound the problem.
Regarding data, any empirical research project must be able to
answer a basic question: When is there enough data? How do we
know when to stop with the data collection? If done systematically,
process tracing can help provide an answer, with my work on socialization in European institutions providing a case in point (Checkel
2001, 2003). After two rounds of interviewing, I took a break from
data collection. Writing up the results – connecting the data to the
causal-process story I was attempting to tell – allowed me to see where
my data coverage was still weak. This indicated in a very specific
manner the data I would need to collect during future field work. This
strategy is consistent with the ethnographic and Bayesian perspectives
discussed earlier, which argue that one should stop when new data
are simply providing additional evidence for the operation of the
hypothesized causal mechanisms.
Transnational dynamics of civil war
23
On method and procedure, good process tracing requires explicit
attention to triangulation (more precisely, its limits) and structural
context. On the former, there is an assumption that triangulation
is a form of methodological nirvana, minimizing threats to causal
inference. With triangulation, a researcher cross-checks the causal
inferences derived from his/her process tracing by drawing upon
distinct data streams (interviews, media reports, documents, say).
The belief is that the error term in each stream points in such a way
that it cancels those in others. However, what if the errors cumulate,
with the result being that the researcher is worse off after triangulating than before (see also Kuehn and Rohlfing 2009)? Indeed, what
if there is a “dark side of triangulation” (Symposium 2007a, 10)?
Well-executed process tracing thus requires that one think beyond
triangulation – for example, by making use of counterfactuals – when
seeking to strengthen the validity of causal inferences.
Another ‘dark side’ to process tracing is how use of the technique
can too easily blind a researcher to broader structural context
(see also Bennett, this volume). For example, in earlier work,
I used a causal-mechanism/process-tracing toolkit to explore the
social-psychological and institutional factors that might lead
decision-makers to change their minds in light of persuasive appeals
(Checkel 2003). Yet, as critics noted, I had overlooked structural
context, simply assuming that persuasive arguments were a function
of individual-level dynamics alone. It was equally plausible, however, that my persuader’s arguments were legitimated by the broader
social discourse in which he/she was embedded. In positivistempiricist terms, I had a potential problem of omitted variable
bias, while, for interpretivists, the issue was one of missing the
broader forces that enable and make possible human agency
(Neumann 2008; Autesserre 2009). In either case, the lesson for the
process tracer is clear: his or her design must incorporate explicit
checks for this potentially serious problem.
In sum, good process tracing builds in part on fairly standard
injunctions and checks that are applicable to an array of qualitative
methods. These include attention to design, thinking counterfactually,
and being cautious in the application of triangulation. At the same
time, it demands adherence to additional best practice standards that
address problems related to equifinality and issues of infinite regress.
Put differently, process tracing, when correctly executed, is far more
24
Jeffrey T. Checkel
than a temporal sequencing of events or “detective work” based on
hunches and intuition (Gerring 2007a, 178).
Extensions to violent conflict
Is any of this useful to students of civil war? My examples above often
derive from situations where institutions work, social relations are
stable, and norms are largely shared; in many cases of civil conflict,
none of these conditions hold. Yet, I will argue that a decisive turn to
the theory and process-tracing practice of causal mechanisms is an
essential next step in the civil-war research program. I substantiate this
claim in two steps, addressing both feasibility and necessity.
Regarding feasibility, is a mechanism/process account – with its high
demands in terms of data, time, and proximity to the subject matter –
realistic in a conflict situation? In fact, a growing array of primary and
secondary source material indicates the data problem is not as severe as
some might expect. Researchers can now draw upon a growing civil
war memoir literature (of former child soldiers – Beah 2007a, 2007b)
and carefully designed surveys of ex-combatants (Blattman 2007;
Annan et al. 2009). In addition, more and more scholars are addressing
the proximity challenge by going directly to the source, conducting
extensive interviews with ex-combatants as a way of measuring causal
mechanisms (Wood 2003; Checkel 2008b; Autesserre 2009; Cohen
2010; see also McAdam, Tarrow, and Tilly 2008, 10–11).
A nagging question remains, though. While there may be more data
available on processes and mechanisms, even in conflict situations, is it
reliable? In such settings, individuals have multiple reasons to lie or to
otherwise dissimulate. Two techniques, however, allow one to bound
this problem. First, just as in non-conflict situations, one triangulates
(cautiously – see above) across multiple data streams. On the particular
issue of interviewees and their possible dissimulation, one looks for
changes and discrepancies in accounts as a function of different audiences addressed. Second, the silences, gestures, jokes, and apparent
lies of interviewees should be recognized as valuable data in their
own right, helping researchers better appreciate how the current
social-political landscape shapes what people are willing to say (Fujii
2010; see also Wood 2003, 33–40).
Of course, in conflict/post-conflict situations, it is not just individuals who may have incentives to lie. Organizations as well may
Transnational dynamics of civil war
25
intentionally distort historical events in a manner designed to meet
current (political) needs, or simply lack the resources to get the story
right. However, here too, good research practices suggest ways to deal
with such realities. In her work on the Peruvian civil war (1980–2000),
Leiby sought to document the frequency of wartime sexual violence,
and the Peruvian Truth and Reconciliation Commission (TRC) was a
key data source in this regard. Finding serious gaps and omissions in its
work, she worked systematically to correct them by consulting other
published documents, as well as primary sources – recoding a random
sample of 2,000 of the 17,000 witness files compiled by the TRC. The
end result was a much richer account of the nature and frequency of
sexual violence in wartime Peru (Leiby 2009a; see also Leiby 2009b).
This discussion of feasibility begs a different and prior question.
Do we even need to bother with adding mechanisms and process to
our accounts? Here, the consensus answer among all students of civil
war – quantitative (Sambanis 2004; Blattman 2007, 24–27; Gates
2008; Blattman and Miguel 2010) and qualitative (Tarrow 2007;
Kalyvas 2008; Wood 2008) – is yes. For very understandable reasons,
one particular social process of civil war – violence – has received
significant attention (Kalyvas 2006). However, as researchers have
moved to disaggregate in recent years – creating subnational data sets,
exploring transnationalism’s role, looking inside rebel groups – new
processes have been invoked, including community building and
allegiance (Gates 2002, 2011), transnational ethnic kinship (Cederman,
Girardin, and Gleditsch 2009), socialization (Cohen 2010), norm-driven
behavior (Wood 2008), emotional responses (Wood 2003), framing
(Autesserre 2009), and social networks (Tarrow 2007), among others.
This fact suggests that not only is research on the processes and mechanisms of civil war feasible, it is also necessary and needed.
Conclusions and preview
This chapter and the chapters that follow argue for improved understanding of civil war and its transnational dimensions by integrating
insights, concepts, and methods from heretofore disparate literatures
and research communities. Dialogue of this type is no guarantee of
theoretical progress or mutual learning; however, its continued
absence is a sure-fire way to promote group think, closed citation
cartels, and academic hyper-specialization.
26
Jeffrey T. Checkel
If one recent state-of-the-art review of the civil war literature can be
read as a plea to reconnect it to central concerns of comparative
politics (Tarrow 2007, 589, 596; see also King 2004, 432–433), then
our argument is that similar gains are to be had via a reintegration with
contemporary international relations theory and methods scholarship.
As two of the sharpest (quantitative) scholars of civil war have recently
argued, the “empirical salience of . . . international issues in driving
domestic civil conflicts” will require that future work “draw heavily
on the existing international relations literature” (Blattman and
Miguel 2010, 30).
Preview
The book has three parts, with this introductory, framework chapter
comprising Part I. The six chapters in Part II form the book’s core; they
utilize process tracing and other techniques to document the causal
mechanisms of transnationalized civil conflict. Each chapter follows
a similar template, where it: highlights a puzzle linked to the transnational
dimensions of civil conflict; advances an argument incorporating one
or more of the causal mechanisms articulated in this introduction;
addresses questions of methods and data; and – most important – shows
these mechanisms in action in one or more cases. This similarity in
structure facilitates learning and cumulation, both within the book and
among readers.
In Part III, we step back and use our results to address two broader
issues. In Chapter 8, Andrew Bennett probes the research frontier
regarding causal mechanisms. In particular, he systematizes our findings (Part II) by advancing a taxonomy of theories about mechanisms,
followed by a discussion of so-called typological theorizing as a means
of addressing combinations of mechanisms. Drawing upon the book’s
case studies, Bennett argues that explanation via causal mechanisms
entails important costs, most notably a substantial loss of parsimony,
but that this is the necessary price for improving our understanding of
the complexities of civil conflict.
In the book’s final chapter, Elisabeth Wood explores how our
approach and findings contribute to the theoretical and methodological
development of work on civil war and its transnational dimensions.
Recent theory on the diffusion of policy serves as her baseline for
assessing our contributions to the understanding of transnationalized
Transnational dynamics of civil war
27
civil conflict. While Wood finds much to praise, she also critiques our
underspecification of the politics at work in any diffusion process,
conflict-driven or otherwise. On method, she reinforces the message
of both this chapter and Bennett’s on the central importance of process
tracing as a technique to identify the observable implications of causal
mechanisms in conflict situations.
PART II
Transnationalized civil war
|
2
Copying and learning from outsiders?
Assessing diffusion from transnational
insurgents in the Chechen wars
kristin m. bakke
Ever since September 11, 2001, Russian leaders have eagerly claimed
that their struggle against rebel forces in the country’s Chechen Republic
is part of a global war on terror, linked to international terrorism.
Similarly, in Moscow during summer 2005, a young Chechen man
explained to me that the second war in Chechnya, which began in
1999, was caused by Moscow and Wahhabis – the name many Chechens
use for people with a radical Islamist leaning – and was funded by
Arab money.1 These examples from Russia demonstrate that both state
leaders and the man-in-the-street have come to emphasize the transnational influences on a struggle that started out largely as a domestic
one. Indeed, in June 2010, the US Department of State (2010) noted that
the activities of the Chechen insurgent leader Doku Umarov, “illustrate
the global nature of the terrorist problem we fight today.” Yet,
while policy-makers, public debate, and a growing body of research have
begun to call attention to such cross-border dimensions of intrastate
struggles, few studies have examined the ways in which transnational
insurgents matter, once they have entered a conflict.
The causes or catalysts for intrastate struggles sometimes rest
abroad. Scholars have argued, for example, that neighboring states
may become sanctuaries for rebel groups (Salehyan and Gleditsch
2006; Salehyan 2007), that nationalist movements may learn or imitate
movements elsewhere (Beissinger 2002), and that diaspora communities
in either near or far-away countries sometimes fund and support rebellions back home (Adamson 2004; Lyons 2006; Smith and Stares 2007).
What this research points to, is that intrastate struggles may not be so
intrastate after all, as domestic challengers to the state are helped or face
pressures from abroad. In this chapter, I explore the domestic dynamics
of the transnational relations of intrastate conflicts. I aim to explain how
transnational insurgents influence the domestic challengers to the state.
1
Personal communication, Moscow, June 2005.
31
32
Kristin M. Bakke
Drawing on the literatures on intrastate conflicts, social movements,
and transnationalism, I theorize the domestic processes that transnational insurgents are likely to impact and the mechanisms through
which they affect these processes.
I begin by briefly describing transnational insurgents. I then situate
the study of transnational insurgents in the literature and present a
theoretical framework for studying how they influence intrastate
struggles. I organize the framework around the domestic insurgents’
mobilization processes likely to be affected by transnational insurgents,
emphasizing how information and resources may travel through
different kinds of diffusion mechanisms. I test the argument through
process tracing, in a study of the Chechen civil wars.
Transnational insurgents
Typically, the term “transnational” describes cross-border contacts
and interactions that are not controlled by states but rather engage
both state and non-state actors (Nye and Keohane 1971, 331). This
study focuses on transnational insurgents in intrastate conflicts, by
which I refer to armed non-state actors who, for either ideational or
material reasons, choose to fight in an intrastate conflict outside their
own home country, siding with the challenger to the state (Malet
2007). Transnational insurgents, who are also referred to as foreign
fighters, exclude foreign legions and private security firms.
The most extensive data collection effort that traces the whereabouts of transnational insurgents is that of David Malet (2007;
2010; 2011), who shows that of 331 intrastate conflicts between
1816 and 2005, transnational insurgents were present in at least 70.
While contemporary policy discussions emphasize so-called Islamist
militants who travel from conflict to conflict (see Hegghammer 2010),
transnational insurgents can also have ethnic ties or other ideological
attachments to the domestic struggle. More than 40 percent of the
conflicts featuring transnational insurgents began after the Cold
War’s demise, in Africa, the post-communist countries, Asia, and
the Middle East. Historically, transnational insurgents have also
participated – or still participate – in conflicts in the Americas (as in
Colombia and Mexico) and Europe (for instance, Greece and Spain).
Clearly, understanding the role of transnational insurgents in intrastate conflicts is an important empirical question.
Copying and learning from outsiders?
33
Transnationalism, social movements, and intrastate struggles
The finding that intrastate conflicts are not so intrastate after all fits
into a long-standing but recently revived research program on transnational relations (see Checkel, this volume, for details). This literature
has primarily focused on peaceful non-state actors, such as
multinational corporations, international organizations, epistemic
communities, and activist networks’ effect on domestic politics (e.g.
Nye and Keohane 1971; Haas 1992; Evangelista 1995; Risse-Kappen
1995b; Keck and Sikkink 1998; Checkel 1999; Tarrow 2005). Thus,
the conflict literature’s discovery that there are violent aspects to
transnational relations is an important contribution.
The emerging literature on violent transnational relations, as well as
the more long-standing literatures on peaceful transnational relations
and social movements, has pointed to a number of variables that
enable non-state actors, including transnational ones, to access or
influence domestic politics. Beginning with the regional context,
conflict-ridden neighborhoods may increase a state’s chances of experiencing violent conflict (Gleditsch 2002; Salehyan 2007). Features of
the state itself matter as well. In particular, so-called failed or weak
states may attract actors with not-so-noble intentions (Rotberg 2004;
Staniland 2005). Similarly, centralized political systems, where
executive power is concentrated, may provide transnational actors
with few access points to their target state, while states where power is
dispersed may provide more access points for influence (Risse-Kappen
1995b). However, states that block domestic groups from exerting
influence may inadvertently force these groups to seek allies outside
the state, as such opening the door for transnational actors (Evangelista
1995; Keck and Sikkink 1998; Pickvance 1999). Thus, transnational
actors may have a harder or easier time gaining access to centralized and
state-dominated polities than to states where societal groups flourish.
At the group-level, features of both the transnational and domestic
non-state actors can affect whether outsiders enter and influence a
domestic struggle. An outside group’s access to policy-makers and
resources will likely aid its impact (Gamson 1975; McCarthy and Zald
1977; Tarrow 1998). Similarly, ethnic affinity or ideological bonds
between foreign and domestic fighters – or the manipulation of such
ties – can increase the chances of outsiders getting involved (Saideman
2002). The domestic fighters, for their part, can vary in the willingness
34
Kristin M. Bakke
and ability to allow transnational actors access. Domestic rebels may be
more willing to welcome the resources that outside insurgents bring
when they face an unfavorable balance of power vis-à-vis the state.2
Some domestic resistance movements may also be less able than others
to keep foreign fighters out. In movements characterized by factional
competition, the actors are not just fighting the state, but also one
another, attempting to eliminate rivals (Pearlman 2009; Lawrence
2010; Cunningham et al. 2012). Transnational insurgents can help
boost the resources needed to fight this dual struggle that each faction
of the domestic movement is waging, and there are multiple entry
points, or domestic allies, for the outsiders. Indeed, while a cohesive
domestic movement may be willing to let outsiders in, it is also able to
say no. A fragmented movement, in contrast, may be less able to control
whether outsiders join its ranks. In the only study that systematically has
begun to examine both when and how transnational insurgents matter,
Staniland and Zukerman argue that both an unfavorable balance of
power and a fragmented domestic movement are necessary conditions
for the emergence of coalitions between domestic and foreign fighters.3
While existing work points to a number of conditions that may
enable transnational insurgents to enter domestic struggles, less work
has focused on the ways in which these actors influence domestic
politics. As Checkel notes in the introduction to this volume, the goal
in statistical studies of the transnational aspects of civil war is typically
to examine whether – to take one example – location in a conflict-ridden
neighborhood leads to violent conflict; however, such correlational
analysis masks the mechanisms at work.
Diffusion from transnational to domestic insurgents
My focus in this study is the mechanisms that link the presence of
transnational insurgents to violent conflict. Consistent with Staniland
and Zukerman,4 I propose that to understand how these outside actors
affect intrastate struggles, one must begin by considering what it is
2
3
Paul Staniland and Sarah Zukerman. “The Effects of Foreign Fighters on Civil
Wars: War Fighting, Ideas, and Recruitment.” Unpublished paper, Department
of Political Science and Security Studies Program, MIT. 2007. They argue that
foreign fighters may lead to changes in the domestic rebels’ ideational motivation,
non fighting capacity, and recruitment.
4
Ibid.
Ibid.
Copying and learning from outsiders?
35
about the domestic challengers that outsiders can shape. Research on
social movements suggests there are three aspects of the domestic
movement that may change as a result of interaction with or pressure
from an outside group – its goals, its repertoire of forms of (collective)
action, and its resources. In military terms, we can think of these as
strategy, tactics, and logistics. They are features that matter for the
domestic movement’s ability to fight the state, either by affecting the
movement’s legitimacy and support in the eyes of the population in
whose name it is fighting, or by affecting its capacity and competence
to organize collectively and confront the state. Indeed, they are closely
linked to key processes in a movement’s mobilization: framing of goals,
tactical innovation, and resource mobilization (see also Schmitz, this
volume; Adamson, this volume). The transnational actors can affect
shifts in these processes through mechanisms such as mediated and
relational diffusion, in turn enabling learning and emulation.
In studies of contentious politics, the term ‘diffusion’ is often used
to describe how social movements may spread from one locality to
another (e.g. Beissinger 2002). However, and as Checkel argues in
Chapter 1, it may be useful to think about diffusion as a process consisting of a number of smaller mechanisms. Initially, there is the mechanism
that enables the spread of information or resources, which can take place
through relational, non-relational, or mediated diffusion (Tarrow 2005,
building on Sageman 2004). Relational diffusion is about the transfer of
information or resources through personal networks and social bonds.
It includes interpersonal interactions. Non-relational diffusion is about
the transmissions between people or groups with no direct ties or social
bonds, where people in one locale learn from people elsewhere through
television, the radio, newspapers, and the internet. Here, I will not
emphasize non-relational diffusion, as my starting point is the presence
of transnational actors. Finally, mediated diffusion takes place when a
third party brings two previously unconnected parties together – or at
least brings together information or resources from two previously
unconnected parties. Mediated diffusion is also called brokerage, as the
third party functions as a broker that brings seller and buyer together.
The parties that are brought together can be communities or movements,
even individuals, and the broker is typically an individual or group of
individuals who may set up institutions. Through either diffusion route –
relational or mediated – learning or emulation on the part of the domestic
insurgents can be based on ideational or instrumental motives.
36
Kristin M. Bakke
Yet transmission of information or resources through these initial
diffusion mechanisms does not necessarily mean that they have any
effect. The effect would depend on subsequent mechanisms –
whether the new ideas resonate with the local population (Keck
and Sikkink 1998; Checkel 1999), and whether people actually
adopt or adapt to the new ideas or use of new resources (Acharya
2004). If transnational insurgents provoke a change in framing or
tactics, one might expect a counter-reaction from the local population, pending on whether the new frame or new tactics match
local views on what is acceptable and useful. The counter-reaction
can deepen existing cleavages within the domestic movement, even
bringing about new divisions. How such counter-reactions occur is a
separate research question.5 The important point here is that the
initial diffusion mechanisms can ultimately have either helpful or
harmful effects on the domestic movement’s mobilization – or no
effect at all.
Shifts in framing
The social movement literature has long emphasized framing as a
process that affects a movement’s ability to mobilize supporters.
Framing implies that the actors define what they are fighting for and
who they are fighting against, often in binary us-versus-them terms.
It includes mechanisms such as the attribution of threat or, in more
clinical terms, diagnosis of the ills that need to be cured and prognosis
for the solution, including the (re)stating and (re)imagining of a legitimate purpose (Snow and Benford 1992; Benford 1993; McAdam et al.
2001, 48).
Transnational insurgents can contribute to shifts in the domestic
movement’s framing through both mediated and relational diffusion
mechanisms that may engender learning or emulation of new frames.
Staniland and Zukerman suggest two routes that correspond to
these types of diffusion: in the long-run, transnational insurgents can
set up schools or other institutions, such as mosques, that transmit
their ideologies to (future) domestic fighters through learning, even
5
Kristin M. Bakke. “Acceptance and Resistance to Foreign Ideas: Transnational
Insurgents’ Impact on the Chechen Separatists.” Unpublished manuscript.
Department of Political Science, University College London. 2010.
Copying and learning from outsiders?
37
indoctrination.6 In the short-term, the domestic fighters’ personal contact
with the transnational insurgents can cause a more direct shift in
goals, especially if the transnational insurgents are successful or possess
charismatic leadership, and if the bond that develops is a close one.
In both routes, the mechanism is about diffusion through learning or
emulation. The diffusion described in the long-term route is mediated
diffusion, or brokerage, where a third party brings together previously
unconnected actors (Tarrow 2005, 104). The more direct diffusion
described in the short-term route is relational diffusion, where learning or emulation results from interpersonal interaction. Studying
recruitment into cults and sects, Stark and Bainbridge note that strong
personal bonds mean that, “[r]ather than being drawn to the group
because of the appeal of its ideology, people were drawn to the
ideology because of their ties to the group” (1980, 1379). Learning
or emulation on the part of the (future) domestic insurgents can be
either ideationally or instrumentally motivated. They may believe in
the new framing, or they may decide to adopt or adapt to this framing
if it is effective at garnering support or resources.
While it may be impossible to empirically establish whether cognitive
mechanisms such as learning or emulation take place, we can observe
the first and last steps of diffusion (but see Nome and Weidmann, this
volume). We can observe whether people are connected either directly
or indirectly through a third party, and we can observe the consequences of diffusion, namely a shift in framing. Hence, if transnational
insurgents affect shifts in a domestic movement’s frame, we would see a
change in framing taking place after the transnational insurgents enter
the struggle – timing is key here. If the mechanism is about mediated
diffusion and long-term learning or emulation, we would observe
the establishment of institutions, schools for example, preceding the
shift in framing. While it is more difficult to establish the presence of
relational diffusion, accounts of heroic individuals among the transnational insurgents, followed by a shift in the domestic movement’s
framing, would suggest that such a mechanism is at work. Importantly,
we should also consider alternative reasons for a shift in framing, as
the change may be homegrown rather than imported.
6
Paul Staniland and Sarah Zukerman. “The Effects of Foreign Fighters on Civil
Wars: War Fighting, Ideas, and Recruitment.” Unpublished paper, Department
of Political Science and Security Studies Program, MIT. 2007. 7 8.
38
Kristin M. Bakke
To the degree that these diffusion mechanisms cause a shift in the
domestic movement’s framing of the struggle, we should carefully
consider the effects on public support. If the new framing resonates with
the local population, we may observe that initially hesitant factions
of the domestic movement adopt the new framing as a means to ensure
public support. Yet the transnational insurgents may fail to strengthen
the domestic movement if they foster a new framing of the struggle
that fails to resonate with the locals.
Tactical innovation
Just like framing is key to a movement’s ability to mobilize supporters
and fight the state, so is its repertoire of forms of (collective) action – its
repertoire of tactics. Tactical innovation refers to a movement’s creation
of new tactical forms (McAdam 1983, 736). A long-standing claim
in the social movement literature says that movements adopting noninstitutionalized or radical forms of contention are more likely to have
an impact than movements that operate within the bounds of “normal”
politics (Gamson 1975; Piven and Cloward 1979). Yet it is also plausible that movements that adhere to the norms of “normal” politics are
the ones most likely to have an impact. A movement’s turn to violence,
for instance, may backfire if that strategy alienates the very population
whose support it needs.
Setting aside the question of whether radical forms of contention help
or hurt a movement’s ability to fight the state, the question here is how
transnational insurgents may affect a domestic movement’s tactical
innovation. In particular, I want to explore their effect on a movement’s
use of radical tactics. By radical tactics, I refer to tactics that the
international community considers inappropriate wartime conduct,
including the intentional killing of civilians, torture, hostage-taking,
and extrajudicial executions. Do transnational insurgents encourage
tactical innovation among the domestic fighters, towards explicitly
targeting non-combatants and resorting to hostage-taking and torture?
Similar diffusion mechanisms as those affecting framing can influence tactical innovation. Mediated diffusion, or brokerage, through
institutions or third parties can foster learning or emulation of new ideas
about morally accepted or effective and efficient tactics. Relational
diffusion through one-on-one interactions with the transnational insurgents, either by fighting side-by-side or in training camps, can engender
Copying and learning from outsiders?
39
learning or emulation of new tactics, especially if those tactics have
proven successful elsewhere and do not contradict local norms for
acceptable behavior. As for observable implications, we would have
to carefully examine whether the transnational insurgents do indeed
advocate or use radical tactics, as well as whether tactical innovation
in the domestic movement, towards more radical tactics, takes place
after the transnational insurgents enter the struggle, keeping in mind
that, alternatively, the sources for innovation may come from within
(Giugni 1999).
Again, we should not assume that tactical innovation enhances the
domestic movement’s ability to fight the state. Tactical innovation can
backfire if it fails to resonate with local norms for appropriate behavior
or local assessments of useful tactics.
Resource mobilization
Perhaps the most common assumption about the traits of a resistance
movement that matter for its ability to fight the state is the movement’s
material resources – its ability to mobilize resources and use these
resources to mobilize people (McCarthy and Zald 1977). Resources
include fighters, weapons, communication, know-how, and finance.
In the literature on revolutions, scholars have long pointed to the
importance of overcoming the collective action problem (Popkin
1988; Lichbach 1994). It is reasonable to expect more resource-rich
movements to be better able to distribute selective incentives that can
lure participants to engage in collective action, even risk their lives,
than resource-poor movements that can offer few, if any, rewards.
Besides helping prevent free-riding, resources, especially coercive
resources such as fighters and weapons (and knowing how to use those
weapons skillfully), are critical in and of themselves: the more coercive
resources the movement possesses, the better able it is to fight the state.
To the degree that transnational insurgents join forces with the
domestic fighters, they may bring fighters and weapons – or funds to
buy those – that increase the coercive strength of the domestic movement. That is, they may directly transmit resources to the domestic
insurgents by bringing competence and capacity. If they do, we should
be able to observe an increase in resources following the entry of
the transnational insurgents. More indirectly, through mediated diffusion or brokerage, transnational insurgents can link the domestic
40
Kristin M. Bakke
movement with funding sources elsewhere, as such contributing to
the movement’s resource mobilization. While it is unlikely that we
can observe the movement’s cash flow by examining its budget, the
evidence for this latter mechanism could be in the form of expert
assessments of a movement’s funding or statements from the domestic
fighters. Again, a consideration of timing would be key to determine
whether the link to funding outside the country is established only after
the transnational insurgents enter the stage.
Unlike transnational insurgents’ effects on goals and tactics, their
effect on resources is in the short-run likely to go just one way: the
more resources transnational insurgents bring, the better able the
domestic challengers are to fight the state. However, if the domestic
movement becomes dependent on external sources of funding, the
long-run consequences can be dire if the funding sources dry up. Akin
to arguments about foreign aid stifling local initiative in developing
countries, the long-run consequence of external sources of funding is
potentially a vulnerable or dependent domestic movement.
In sum, assessing the role of transnational insurgents requires establishing the following about the domestic movement: Do we observe a
change in framing of its goals? Does it adopt new tactics and carry out
operations that were previously taboo? Do the local fighters have more
resources or carry out operations for which they previously lacked the
resources? If we do observe these changes, we can not conclude that
they are caused by transnational insurgents until we have traced the
changes back to mediated or relational diffusion from these outsiders,
as well as considered alternative domestic reasons for change.
Research design
The research task in this study requires a process-tracing approach
(Checkel, this volume; George and Bennett 2005; Gerring 2007a). As
I am interested in tracing the mechanisms that create variation across
three processes of domestic mobilization – framing, tactical innovation,
resource mobilization – the case selection must allow for such variation
(on the dependent variables). To that end, I examine Chechnya’s
conflict with the Russian federal government over time, which offers
shifts in each mobilization process. The conflict started with
the Chechen declaration of independence in 1991, turned into a
war in 1994, came to an end with a ceasefire in 1996, again turned
Copying and learning from outsiders?
41
violent in 1999, and today has come to an uneasy stalemate,
with violence dwindling since 2005.
In terms of framing, the Chechen conflict began as a nationalist
struggle much like the rest of the “parade of sovereignties” in Russia
in the early 1990s. It remained so throughout the first war, focusing on
independence for the Chechen republic. Since the period between the
first and second wars, Chechen leaders increasingly made references to
the establishment of an Islamic state. In recent years, some Chechen
resistance leaders have framed the struggle in terms of nationalism and
self-determination, while the movement’s dominant branch is fighting
an Islamist struggle aimed at creating an Islamic emirate.
As for tactical innovation, one of the infamous characteristics of the
second Chechen war was a growing kidnapping-for-ransom industry,
which did not to a similar extent characterize the insurgents’ tactics in
the first war – although both the Chechens and the Russians were
known to have carried out kidnappings even in the first war. Another
infamous characteristic associated more with the second war than the
first was large-scale terrorist attacks outside Chechnya’s borders, such
as the Dubrovka/Nord-Ost theater siege in Moscow in 2002 and the
Beslan school siege in 2004 – although again, attacks had taken place
in 1994–1996. Moreover, in 2000 suicide terrorism became a new tool
in the Chechen insurgents’ repertoire of tactics. These trends suggest
that the tactics of the Chechen resistance movement have changed over
time, turning more radical in the sense that also civilians have become
explicit targets of violence.
In terms of the ability to mobilize resources and muster coercive
strength, the Chechen struggle has also changed over time. In the
immediate post-Soviet era, the Chechen separatist government
inherited weapons from the branch of the Soviet army that had
been stationed in Chechnya and received funding primarily from the
Chechen community in Russia. In the second war, the Chechens
turned to home-made weapons, but they reportedly used them with
great skill (Dudayev 2004). Due to the destruction of the first war, the
relative popularity of Putin’s war on Chechnya within Russia, and the
fact that the resistance leaders from 2000 onwards did not officially
head their republic (as Putin put Chechnya under central control), the
Chechen fighters had fewer domestic resources available than in
the first war. Reports indicate, however, that resources from outside
increased.
42
Kristin M. Bakke
To assess if and how transnational insurgents have affected these
mobilization processes, I consult a variety of sources, including personal
communication with Chechens living in Moscow; the documentary/
propaganda film The Life and Times of Khattab, which contains interviews and footage of the most important transnational insurgent in
Chechnya (Waislamah News Network 2002);7 the Islamist website
Kavkaz Center; the separatist website Chechenpress; monitoring of local
Chechen papers by the Russian International Institute for Humanities
and Political Studies (IGPI); journalistic accounts through the Jamestown
Foundation (Chechnya Weekly, Terrorism Monitor, North Caucasus
Analysis, Prism), which is an independent institute with extensive coverage of the North Caucasus; reports from the Institute of War and Peace
Reporting (Caucasus Reporting Service); news searches via Lexis-Nexis;
and academic and biographical articles and books.
Transnational insurgents in the Chechen wars
While it is impossible to get precise data on the number of transnational insurgents in Chechnya at any given point – and the Russian
authorities may have sought to inflate the numbers – it has been
suggested that over the course of the two wars, 500–700 transnational
insurgents, including members of the diaspora, have fought there
(Moore 2007). A total of 500–700 indicates an increase over time,
from the 80–90 who were reported to be active in the first war. Some
estimate there were 100–200 transnational insurgents present during
the second war (B. Williams 2005b), which would suggest that the
highest number of transnational insurgents entered in the interwar
period and early years of the second war. The first transnational
insurgents came to Chechnya in February 1995, just a couple of
months into the first war, as followers of the fighter known as Emir
Khattab or Ibn Al-Khattab (Gall and de Waal 1998, 308; Tishkov
2004, 172; Gammer 2006, 214–215). Of Saudi Arabian or Jordanian
origin, Khattab fought in the civil wars in Afghanistan and Tajikistan
before entering Chechnya in 1995, upon the invitation of a member
of the Jordanian Chechen diaspora community known as Sheikh Ali
7
It is a 2002 Arabic language production (with English subtitles) that portrays Khattab
as a war hero, giving it a propaganda feel, using interviews and footage of Khattab.
Produced after Khattab’s death, the film is a collage of earlier footage. It is presented by
Waislamah News Network as Series 1 of Contemporary Heroes of Islam.
Copying and learning from outsiders?
43
Fathi al-Shishani, who had fought against the Soviets in Afghanistan
and had moved to Chechnya in 1993 (Moore and Tumelty 2009, 84).
According to some accounts, Khattab brought with him a unit of eight
Afghan-Arab fighters. In his own words in The Life and Times of
Khattab, Khattab says he entered Chechnya with 12 “brothers” from
Dagestan, then meeting “other groups,” although it is not clear who
these other groups were. Tumelty (Moore and Tumelty 2009) reports that
after Khattab’s arrival in Chechnya, he met with Fathi, who had recruited
some 90 Arab fighters from Afghanistan, of which 60 joined Khattab. The
transnational insurgents active in Chechnya have primarily been from the
Middle East, but some have also come from North Africa, Turkey, and
possibly Pakistan. The question here is if and how these transnational actors
over time have affected the Chechen resistance movement’s framing of
goals, tactical innovation, and resource mobilization.
Effect on framing
The Chechen insurgency grew out of the Chechen nationalist movement
that emerged in the final days of the Soviet Union. In fall 1991, under the
leadership of Dzhokhar Dudayev, a Chechen who had served as a Soviet
Air Force general, the nationalists overthrew the local communist-led
local government and declared the republic independent.
Leading up to and during the first war, the nationalists framed
the struggle around an image of an aggressive and exploitative state that
consistently had imposed suffering on them (Radnitz 2006). Dudayev
proclaimed in a 1991 interview that, “I will restore my people’s pride after
our enslavement by the Russians” (Sheehy 1991, 26).8 This sentiment
echoed many Chechens’ collective memory, which was – and still is –
colored by Stalin’s deportation of nearly the entire Chechen population
from its homeland between 1944 and 1957 (B. Williams 2000; Tishkov
2004, 50–54; Ustinova 2004). To many Chechens, the deportation is part
of a repeated history of repression. In 2005, a Chechen man explained to
me that the heart of the Chechen question is that with regular intervals,
Russia has occupied Chechnya. The tsarist forces did it, then
the Bolsheviks, Stalin, and now the current post-Soviet regime.9 Such
views were not uncommon in the early 1990s, and Tishkov (2004, 53)
8
9
See also “Interview with Jokhar Dudayev by Peter Collins, Voice of America,”
Official Kremlin International News Broadcast, December 11, 1995.
Personal communication, Moscow, June 2005.
44
Kristin M. Bakke
argues that Chechens increasingly believed that self-determination was
necessary to halt long-time discrimination.
Dudayev was not opposed to Islam, and he was more open to the
teaching of Islam than his Soviet predecessors; he even used traditional
Sufi Islamic references to mobilize people. The State Islamic Institute
in Grozny (the Chechen capital), which received funding both from
Dudayev’s government and Saudi Arabian and Islamic International
foundations, opened in 1991 (Bobrovnikov 2001, 13). This nod to
Islam must be seen in light of Dudayev coming to power in the early
post-Soviet period, when the Chechens were happy to again freely
practice their religion, long oppressed. Yet Dudayev openly favored
a secular state (Tishkov 2004, 169). He formed a commission to
consider Islamic sharia courts, but the legal system remained secular
until after he was killed in 1996 (Muzaev 1997). Indeed, while there
were voices calling for an Islamic state in the early 1990s (German
2003, 31; Moore and Tumelty 2009, 83), the Chechens’ initial quest
for independence had little to do with religion.
This changed when Dudayev was killed by Russian forces in April
1996. His immediate successor, Zelimkhan Yandarbiyev, had ever since
the founding of the nationalist movement favored an independent and
Islamic Chechen state (Dunlop 1998, 93). In his brief period in power,
Yandarbiyev called for the introduction of sharia criminal code (Powell
1996).10 When asked in an interview whether it would be natural for
Chechens to live under sharia law, Yandarbiyev responded that, “We
are fighting to protect our independence and to defend the honor and
dignity of the free people under the banner of Islam.”11 The struggle was
focused on Chechnya, blending nationalism and Islamism:
Our jihad is first of all a jihad to defend the territory, honor and dignity of the
Chechen people. But a jihad is not only a war conducted with arms in hand.
It is a struggle against everything that contradicts what has been established by
the Single God, that is against everything that a Muslim is forbidden by God.
In this sense, of course, the jihad knows no concrete forms or borders.12
10
11
12
See also “Guard to be set up in Chechnya to Supervise Implementation of Sharia
Law,” BBC Summary of World Broadcasts (ITAR TASS), October 17, 1996.
“Chechen Separatist Leader Yandarbiyev Thinks Peace Process Will Continue,”
BBC Summary of World Broadcasts (NTV), October 23, 1996.
Quoted in “Interview Granted by Chechen Republic President Zelimkhan
Yandarbiyev,” Official Kremlin International News Broadcast (Nezavisimaya
Gazeta), December 20, 1996.
Copying and learning from outsiders?
45
Among Yandarbiyev’s three most serious contenders in the 1997
presidential elections in the republic, all made reference to Islam
but to varying degrees (Radnitz 2006). Aslan Maskhadov, the field
commander who had negotiated the 1996 peace agreement with
Moscow, advocated a secular state, yet viewed Chechnya’s Islamic
customs favorably.13 Shamil Basayev, a prominent field commander,
called for an uncompromising attitude to Chechen independence
and adherence to Islam, but fell short of advocating an Islamic state.14
The third candidate, Movladi Udugov, who had served in both
Dudayev and Yandarbiyev’s governments, favored an independent
Islamic state. In the elections, Maskhadov emerged as the clear victor,
suggesting a popular rejection of the most radical religious agenda.
Over time, the struggle has taken on much more of a religious tone,
departing from a Chechen-centered one. This development appears to
have taken off in the interwar years, and the second war was, in contrast
to the first war, one where Islamist (Salafi) goals played a significant role.
President Maskhadov was killed in 2005, and both of his successors,
Abdul-Halim Sadulayev, who was killed after only a year in power, and
Doku Umarov, have framed the struggle in more Islamist terms. While
Sadulayev’s framing of the struggle contained elements of both nationalism and Islamism,15 Umarov’s framing is more clearly Islamist. In the fall
of 2007, Umarov proclaimed that his struggle was in the name of a unified
Caucasus Emirate, not just self-determination for any one ethnic group:
I am announcing to all Muslims that I am at war against the infidels under
the banner of Allah. This means that I, Amir of the Caucasian Mujahideen,
reject all infidel laws that have been established in this world. I reject all laws
and systems that the infidels have established on the land of the Caucasus.
13
14
15
“Separatist Leader Maskhadov against Creation of Islamic State,” BBC
Summary of World Broadcasts (ITAR TASS), September 17, 1996; “Separatist
Commander Maskhadov Interviewed on Talks with Russia, Fundamentalism,”
BBC Summary of World Broadcasts (Tyden), October 3, 1996.
“Chechen Separatist Chief Admits Arab Involvement in Conflict in North
Caucasus,” BBC Summary of World Broadcasts (ITAR TASS), November 4,
1996; “Interview Granted by Shamil Basayev, a Candidate for the Chechen
Presidency,” Official Kremlin International News Broadcast, December
16, 1996.
“Russia’s Tactics Make Chechen War Spread across Caucasus Rebel
President,” BBC Monitoring (Chechenpress), September 13, 2005. See also the
Jamestown Foundation interview published after Sadulayev was killed,
published in Chechnya Weekly, July 6, 2006.
46
Kristin M. Bakke
I reject and outlaw all names that the infidels use to split the Muslims.
I outlaw all ethnic, territorial and colonial zones named ‘North-Caucasian
republics’, etc . . . We renounce all these names (quoted in Smirnov 2007b).
Indeed, Umarov, who fought in both wars and was a member of
Maskhadov’s government, has stated that, “Mujahideen in the Caucasus do not fight for democracy, they fight for Sharia” (Smirnov
2007d). From considering himself the president of Chechnya (a position
not officially endorsed as the post is held by the pro-Moscow Ramzan
Kadyrov), he has moved to consider himself the emir of the North
Caucasus Emirate. This framing of the Chechen struggle as a religiously
motivated quest aimed at establishing a larger Islamic state differs from
the nationalist framing leading up to the first Chechen war.
The question here is whether the change in the Chechen separatist
movement’s framing can be traced to transnational insurgents. The
empirics suggest yes. Transnational insurgents influenced the framing
of the movement’s goal primarily through Khattab and his initially small
group of fighters, who were dedicated to a social and political Islamic
order (Wilhelmsen 2004; Gammer 2005; Speckhard and Ahkmedova
2006; Moore and Tumelty 2008, 2009; Sokirianskaia 2010). Indeed, to
Khattab, who had fought in both Afghanistan and Tajikistan, the
Chechen struggle was yet another piece in a wider Muslim struggle.
Timing-wise, while Dudayev also embraced some Islamic teaching and
practices, the shift in framing towards a jihadist struggle, particularly
visible in field commander Basayev’s framing, followed the entry of
transnational insurgents with an Islamist agenda. This change happened
through both mediated and relational diffusion.
Initially, Sheikh Fathi functioned as a broker who brought Khattab
to Chechnya, thus connecting Chechen insurgents with fighters from
the Middle East and elsewhere in the region. This is mediated diffusion:
a third party established links between previously unconnected parties.
Yet the degree to which the Chechen fighters were unconnected to
insurgents abroad should not be overstated. Already in 1994, Basayev
had taken 30 of his fighters to a military training camp in Afghanistan
(B. Williams 2003; Hughes 2007, 101–102). Moreover, in 1992,
Basayev had played a critical role assisting the self-determination
movement in Abkhazia in its armed struggle against the Georgian
state. Thus, even without Fathi, the Chechens had established links
to foreign insurgents and insurgencies.
Copying and learning from outsiders?
47
Once in Chechnya, the route through which information was transmitted from Khattab and his followers to the Chechen insurgents was
both relational and mediated. Khattab initially set up military training
camps in 1995, and in 1996, he established a training center, Kavkaz,
consisting of several camps for both military and religious instruction
near the village of Serzhen’-Yurt in the mountainous eastern part of
the republic (B. Williams 2005a; Moore and Tumelty 2008).16 The
Life and Times of Khattab shows footage of Khattab in Chechnya,
his voiceover recounting how he entered the republic and started
training his “brothers.” The camps included training on religion, mine
cleaning, land mining, and the use of weapons. The students, mainly
young men, were primarily Chechens but also Dagestanis, Arabs,
and Muslims from elsewhere in the North Caucasus and Central Asia.
According to one account, they were paid to attend (Baiev 2003, 206).
The teachers were primarily foreigners speaking Arabic, helped
by translators from Dagestan (Sokirianskaia 2010, 212). Hundreds
of young people passed through this training center, and many went
on to become fighters in the Chechen resistance in the second
war (Sokirianskaia 2010, 212; Moore and Tumelty 2009, 85). In
terms of mechanisms, these training camps were sites for both relational diffusion – through direct hands-on training by transnational
insurgents – and mediated diffusion – to the degree that non-insurgent
teachers introduced their students to new ideas.
Khattab’s increasingly close relationship to the field commander
Basayev, who allegedly in spring 1998 appointed Khattab his foreign
security advisor,17 is also an example of relational diffusion. Indeed, in
his online memoir Book of a Mujahiddeen [sic], written in 2004,
Basayev specifically mentions Khattab’s lesson on strategies and tactics
of war.18 In 1998, Khattab and Basayev jointly organized the Islamic
International Peacekeeping Brigade, which consisted of Chechen and
foreign fighters from the Middle East and North Africa (Wilhelmsen
16
17
18
Vidino (2005) notes that many reports of training camps come from Russian
intelligence sources and Islamist propaganda videos, but Sokirianskaia’s (2010)
ethnographic study includes interviews with Chechens who passed through these
camps, providing confidence in the camps’ whereabouts and programs.
Sanobar Shermatova, “‘I Feel Secure When I Have Ammunition at Hand’,”
Moscow News, April 16, 1998.
I was made aware of this e memoir, which I located by searching online, by King
(2008, 241). The online location of this publication changes over time. It was last
accessed at www.tawhed.net/dl.php?i=0512091i (August 15, 2012).
48
Kristin M. Bakke
2004, 34). The two were responsible for the event that triggered the
second war with Russia, the attack into neighboring Dagestan in
August 1999. Unsanctioned by President Maskhadov, the attack was
aimed at creating an Islamic Republic of Chechnya and Dagestan
with the help of Dagestani Islamist militants.
While it is true Basayev had led an attack into neighboring Dagestan
in June 1995 – taking more than a thousand civilians hostage at a
hospital in Budennovsk – that incident was framed as a response to
the Russian forces’ violent campaign in Chechnya. “Our aim was to
reach Moscow . . . and our purpose is to stop this war,” he argued.19 It
was an attack motivated by revenge and Chechen independence. In
contrast, in the 1999 attack into Dagestan, Basayev framed it as a
struggle aimed at driving the Russian “infidels” out of the North
Caucasus region: “Today there is a great deal of work for brother
Muslims from all over the world . . . we will fight until the full victory
of Islam in the world.”20 These words echoed those of Khattab, who
never had seen Chechnya as an isolated struggle. In a 1997 interview,
for instance, Khattab said that, “We are preparing ourselves for other
jihad (holy war) operations in Chechnya or elsewhere.”21 Basayev’s
shift in framing, towards a larger Islamist struggle, followed Khattab’s
entry into Chechnya.
Basayev’s 1999 attack into Dagestan, which was the manifestation
of a shift in framing, was also apparently influenced by the Dagestani
Wahhabi preacher Bagaudin Kebedov, who had fled the crackdown on
radical Islamists in his own republic and entered Chechnya in December
1997 (Muzaev 1998e; 1999b; 1999d). Prior to the August 1999
invasion, he asked Basayev to assist him in overthrowing Dagestan’s
pro-Russian rulers and establish an Islamic Republic (Vatchagaev
19
20
21
Quoted in “Rebels Execute Russian Captives,” St. Petersburg Times (Florida),
June 16, 1995. See also David Zucchino, “War Stories from Rebel on the Run:
Raid Was a Wake up Call for Russians, Chechen Explain,” The Philadelphia
Inquirer, July 16, 1995.
Quoted in Ruslan Musayev, “Prominent Chechen Warlords Lead Rebels in
Fighting Russian Troops,” Associated Press Worldstream, August 11, 1999; see
also “Chechen Says He Leads Revolt in Nearby Area,” New York Times,
August 12, 1999.
Quoted in “Mujahedeen Leader Says Chechnya Still Faces Russian Danger,”
Agence France Presse, May 17, 1997. See also interview in Sanobar Shermatova,
“‘I Feel Secure When I Have Ammunition at Hand’,” Moscow News, April
16, 1998.
Copying and learning from outsiders?
49
2007). Yet while it looks like Kebedov may have played a role in
influencing Basayev’s 1999 Dagestan attack, the timing suggests
that Kebedov followed, rather than fostered, a shift in framing. He
did not enter Chechnya until late 1997, for the very reason that the
increasingly pro-Islamist environment in Chechnya was more suitable
for his preachings than the environment in Dagestan.
Diffusion may also have taken place via the more long-term mediated type, in particular, through schools and mosques that introduced
young people to new ideas. From 1996–97, Salafi mosques and schools
were opened in Grozny and rural regions of the republic, sometimes
with financial rewards to the regions that adhered to this new and nonindigenous version of Islam (Speckhard and Ahkmedova 2006, 16). In
The Life and Times of Khattab, Khattab recounts how these religious
institutions had structured programs, with training for memorizing the
Koran and courses for giving dawah (spreading the message of Islam),
organized from basic to advanced levels. The best source of information
for whether this kind of diffusion had lasting effects is Ekaterina
Sokirianskaia’s (2010) extraordinary Ph.D. dissertation on statebuilding in Chechnya and Ingushetia, based on years of fieldwork. Her
assessment is that, while some of the leaders in the insurgent movement
may have turned to Islam for instrumental reasons, “for young people
who followed these opposition leaders, who went through the training
camps of Khattab, Islamism was already in earnest” (2010, 216).
This turn towards Islam at the societal level made it difficult for
the government under President Maskhadov not to turn to it as
well. While not a proponent of an Islamic state and religious courts,
Maskhadov was gradually compelled to establish sharia courts and
implement Islamic rules across the republic (Muzaev 1998a; 1998c;
1998d; 1998e; 1999c), and in February 1999 he implemented full
sharia law (ibid. 1999a). It is plausible to reason that because
Maskhadov’s turn towards Islam followed the establishment of Salafi
schools and mosques, he may have found that a more religious message
was key to keeping the support of his constituents – who may have
attended these schools or mosques – and who looked favorably upon
the Wahhabis who had come to join the Chechens’ struggle (Gammer
2005). In other words, it was an instrumental attempt at co-optation.
Indeed, the motivation for turning towards a more Islamist struggle
appears to have been driven both by instrumental and ideational
motives. Because there were sometimes financial rewards attached to
50
Kristin M. Bakke
turning towards the new branch of Islam that the outsiders brought
along, including access to financial patronage abroad (e.g. Moore and
Tumelty 2008, 419), one cannot claim that diffusion necessarily took
place through learning or genuine adoption of these ideas; it may rather
have been emulation in response to these rewards. In a 2003 interview,
the Chechen commander known as Amir Ramzan, who admitted to
receiving financial support from the transnational insurgent Abu Walid
al-Ghamdi, said the following when asked if his foreign patrons
expected anything in return: “The most important thing for them is that
the money is used for the war, for the jihad, and that those who receive
it are true Muslims.”22 Similarly, Sokirianskaia quotes Yandarbiyev:
“Islamic fundamentalism is not dangerous. It is partnership, international relations. You do not consider it a problem if Western
investors tour Russia, do you? One cannot divide help into help
from Wahhabis and help from others” (quoted in Sokirianskaia
2010, 215). This comparison of Islamic fundamentalism to investors
suggests that instrumental motives cannot be ruled out.
While the struggle over time has come to include more references to
Islam and establishing an Islamic state, it is not exclusively an Islamist
struggle. Even Basayev, who was killed by Russian security forces in
summer 2006, stated in a 2005 interview with the Russian journalist
Andrei Babitsky, broadcast on ABC Nightline, that religion was not as
important as national liberation: “I need guarantees that tomorrow
future Chechen generations won’t be deported to Siberia, like they
were in 1944. That’s why we need independence.”23 A few months
later, though, in February 2006, a more Islamist frame dominated:
“Today, before every Muslim is the duty to take the path of jihad
and fulfill the command of Allah” (quoted in Jamestown Foundation
2006). Internal communication between Basayev and his soldiers
suggest that he, indeed, was motivated by religion. Yet in public
statements that could reach a Western audience, he may have been
more reluctant to say so (Radnitz 2006, 249), again suggesting that
instrumental concerns were key to the framing of the struggle.
Although the framing of the struggle today is more colored by
references to an Islamic state, sharia law, and the unity of Muslims in
22
23
“Next Year the War Will Seize the Entire Caucasus,” Kavkaz Center, November
28, 2003. Available from www.kavkazcenter.com/eng/content/2003/11/28/
2039.shtml (accessed June 7, 2011).
ABC News Nightline, July 28, 2005.
Copying and learning from outsiders?
51
the North Caucasus than in the first war, it also encompasses the
rhetoric of self-determination. While the Chechen resistance movement
until 2007 had one overarching leader, it is now divided between the
nationalist branch, headed by the UK-based Akhmed Zakayev, who
served as foreign minister under Maskhadov and is now the selfproclaimed prime minister of the Chechen government in exile, and
the Islamist branch under Umarov. So far, this division has primarily
played out as a war of words, but Chechnya’s trajectory in the interwar
period, when the Islamist framing took root, suggests that divisions
within the movement can be detrimental. Indeed, the Chechens’ second
war with the central government in Moscow resulted from Maskhadov’s
inability to rein in different factions, in particular Basayev and Khattab.
While the empirics suggest that a turn towards a more religious
struggle in Chechnya was influenced by transnational insurgents, they
also point to a conditioning variable and an alternative explanation:
domestic gatekeepers and domestic roots.
First, the Chechen resistance leaders have played a key role in
allowing foreigners to have an impact (B. Williams 2007; Moore and
Tumelty 2008; 2009). Under President Dudayev (1991–1996), who
was not a proponent of establishing an Islamic state, the lead foreign
fighter in Chechnya, Khattab, was careful not to challenge the goals
or strategies of the domestic insurgent movement. Indeed, during the
first Chechen war, Khattab was under the command of Chechen field
commanders (e.g. Gammer 2006, 215; Moore 2007; Sokirianskaia
2010, 212). In part, the domestic resistance leaders’ upper hand
vis-à-vis Khattab in the first war is attributed to the war-effort
being funded primarily by the Chechen population in Russia and
the Chechen diaspora community – and under the control of Dudayev
(Wilhelmsen 2004, 41). Under Yandarbiyev’s brief reign (1996–1997),
Khattab could do more to fulfill his vision of an Islamic state in
the North Caucasus as Yandarbiyev already looked more favorably
upon the idea of an Islamic state – and was head of a then war-torn,
cash-starved republic. While Yandarbiyev’s successor, Maskhadov,
was a secularist and not initially a proponent of establishing an Islamic
state, Khattab’s close alliance to Basayev, who increasingly worked in
opposition to Maskhadov, ensured that he had access to Chechen
insurgents and people.
Thus the influence of transnational insurgents has to a certain extent
been controlled by gatekeepers in the Chechen resistance movement.
52
Kristin M. Bakke
Once the movement shifted away from the relatively centralized structure of President’s Dudayev’s National Guard, it became easier for
outsiders to have an influence. Indeed, Maskhadov’s presidency
was plagued by an inability to control the former field commanders,
particularly Basayev. By the fall of 1998, Maskhadov’s forces were
even engaged in combat with some of the former field commanders
(Tishkov 2004; Gammer 2006). This kind of environment, paired with
the destruction and poverty of the Chechen republic after the first war,
provided space for the influence of outside forces.24 Notably, the
Wahhabi preacher Kebedov escaped from crackdowns on Wahhabism
in Dagestan to Chechnya precisely in this time period. At the same time,
the influence of the transnationals contributed to cleavages and clashes
among different armed factions – in turn hampering negotiations with
Moscow. Aware of the detrimental effects of such infighting, in April
1999, Maskhadov tried, unsuccessfully, to bridge divisions between
different factions (Muzaev 1999b; 1999e).
Second, the empirical record suggests there may be an alternative
domestic explanation for the shift in framing that has little or nothing to
do with transnational insurgents. Prompted about the influence of Arab
fighters in a 2006 interview, Umarov, the leader of the Islamist branch,
claimed it was exaggerated. Moreover, he pointed out that, “[t]hese
Arabs fight everywhere where there is jihad, not only in Chechnya, to
fulfill their Muslim duty” (quoted in Jamestown Foundation 2006),
suggesting that transnational insurgents came to Chechnya because
there already was a jihad – not the other way around. Indeed, the shift
towards a religious struggle may have been part of a purely domestic
strategy to unify the different ethnic groups across the North Caucasus in
a struggle against the Russian state, drawing on the religion that these
groups shared. In earlier wars in the region, Islam had also been the basis
for unity among the different ethnic groups (Gammer 2006, 64–66;
Smirnov 2007c; 2007d), and Islamic rhetoric has a history in struggles
in the region. Writing about the 1990s, the historian Charles King argues:
Web sites such as kavkazcenter.com – a major channel of communication
with the rest of the world – consistently appropriated the Islamic lexicon
found in [Caucasian rebel leader] Shamil’s letters written 150 years earlier.
24
See Paul Staniland and Sarah Zukerman. “The Effects of Foreign Fighters on
Civil Wars: War Fighting, Ideas, and Recruitment.” Unpublished paper,
Department of Political Science and Security Studies Program, MIT. 2007.
Copying and learning from outsiders?
53
Rather than “Russians,” the Web masters spoke of “unbelievers” (kafirs).
Rather than pro-Russian Chechens, they spoke of “hypocrites” (munafiqs).
Casualties on the Chechen side were always “martyrs” (shahids). There is a
world of difference between the Russian Empire’s wars in the Caucasus and
those of the Russian Federation, but in Chechnya of the early twenty-first
century the past still proved a dark template for action (King 2008, 241).
That is, the template for a religious struggle is found both abroad and
in Chechnya’s own history. Across the North Caucasus, especially in
Dagestan, the fall of communism and the centralized Soviet state were
associated with a return to Islamic practices from the late nineteenth
century, including mass pilgrimages to Mecca and Medina (Bobrovnikov
2001). Thus one could argue that the use of the Islamist framing is a
strategic last resort of a weakened resistance movement, hoping to create
a larger Caucasian alliance in its fight against the Russian state. The
timing of a return to such a template, though, during the interwar period,
does suggest that transnational insurgents played a role in its revival.
Effect on tactical innovation
The second Chechen war has become associated with large-scale
hostage-takings, suicide terrorism, and kidnappings. While the
first war also witnessed attacks directed at civilians, notably the
hostage-taking at a hospital in Budennovsk in June 1995 and
the Kizlyar-Pervomayskoye hostage crisis in January 1996 (both in
Dagestan), the second war has been the scene of the most infamous
large-scale terrorist attacks. These include the Dubrovka/Nord-Ost
theater siege in Moscow in October 2002, where more than 800 people
were held hostage (129 killed, including in the rescue operation), and
the Beslan school hostage crisis in September 2004, where more than
1,100 people were held hostage, most of them schoolchildren (a total of
334 killed). Moreover, suicide bombings, which were absent from
the first war, began in the summer of 2000 and peaked between 2003
and 2004 (Nivat 2005; Speckhard and Ahkmedova 2006). While the
suicide attacks initially targeted only military targets, beginning in 2002,
the attacks, several of them featuring women suicide bombers, were also
directed at civilians. Suicide attacks have also come to be employed by
citizens of other North Caucasus regions, including the bombings of
Moscow’s metro in March 2010, under the umbrella of Umarov’s
Islamist struggle. The question for this chapter is whether this tactical
54
Kristin M. Bakke
innovation – towards more radical forms of action – can be traced
to diffusion from transnational insurgents.
The empirics suggest a hesitant yes. Even though suicide attacks did
not take place until 2000, Dudayev had apparently seen such tactics as
an option since 1994 (Hughes 2007). In terms of large-scale hostagetakings, while there has been an Arab presence in all four major
hostage attacks related to the Chechen wars, in each case, those in
charge were Chechen field commanders, and the demands raised were
specific to the Chechen conflict – an end to the conflict, withdrawal of
Russian forces, and independence – and not about a global jihad
(Moore and Tumelty 2008, 426; Speckhard and Ahkmedova 2006, 10).
While the Kremlin played up the Arab presence among the hostagetakers at School No. 1 in Beslan, most of them were Chechen and
Ingush, suggesting local motivations (Tuathail 2009). Similarly, the
turn to suicide terrorism stems from local grievances, and Chechen
resistance leaders have taken responsibility for most of the attacks
(Moore and Tumelty 2008, 427). Yet the fact that Arabs were not
the dominant actors and the demands were Chechen-centered does
not preclude tactics being influenced by transnational insurgents.
Relatively little time passed between the entry of transnational
insurgents in Chechnya, with Khattab in February 1995, and the first
large-scale hostage event, the Budennovsk hospital siege in June 1995.
While Basayev, who orchestrated the attack, early on became a close
ally of Khattab, there is too little information about the degree of their
contact in the early days of 1995 to assess whether Basayev’s choice of
tactics was caused by learning or emulation via relational diffusion.
Basayev claimed that ten Arabs participated in the attack, which
suggests the possibility of relational diffusion, although he also pointed
out that he personally supervised the training of these foreigners – not
the other way around.25 He argued that the Budennovsk attack
was meant to make Russians feel “the real horror of the war
that Moscow had unleashed on his people” (quoted in Quinn-Judge
2004), indicating a homegrown reason for a change in tactics, based
on a violence-begets-violence mechanism. Indeed, many members of
Basayev’s own family, including his wife and two children, had been
25
“Chechen Separatist Chief Admits Arab Involvement in Conflict in North
Caucasus,” BBC Summary of World Broadcasts (ITAR TASS), November
4, 1996.
Copying and learning from outsiders?
55
killed in a Russian air raid on his hometown just two weeks prior to
the attack. Basayev often threatened to again take the struggle to
Russian soil and target civilians,26 but he also claimed that he had
not gone to Budennovsk with the intent of taking civilian hostages
(Gall and de Waal 1998, 260, 263), and that he did not intend
to undertake a similar hostage-taking.27 Indeed, according to some
news reports, the Budennovsk attack was quite spontaneous, almost
accidental – allegedly, the rebels were heading for Moscow but had to
stop as they ran out of money for bribing the road police. Similarly,
the 1996 hostage-takings in Kizlyar and Pervomayskoye, led by Salman
Raduyev, appear to have been carried out as a spontaneous second-best
alternative after the rebels’ attack against a Russian air field in Kizlyar
failed (Jamestown Foundation 1996). After the Budennovsk attack,
Basayev’s deputy, Aslambek Abdulkhadzhiev, emphasized that they
never again planned to carry out anything like it, suggesting that such
an attack on civilians was, somehow, unacceptable:
Here I must say we do not plan anything like Budennovsk. The Budennovsk
tragedy will never be repeated. Moreover, we did not make these plans
except as a last resort. Why was the world was silent when Shali was bombed
[by the Russians], when some 400 people were killed or wounded? (quoted
in Jamestown Foundation 1995)
Abdulkhadzhiev’s assessment here is slightly different than later statements posted on the Kavkaz Center website. Consider the following
2009 statement from Umarov, which seems to indicate a degree of
acceptance of civilian targets:
(T)his year will be also our offensive year all over the territory of Russia.
Why? Because I think that those people who are living today in the territory
of Russia, they are also responsible for their soldiers, for their leadership, for
those atrocities, for that outrage, that they commit, and for those wars that
they wage today against Islam.28
26
27
28
“Next Time It Could be Moscow, Says Chechen Commando Chief,” Agence
France Presse, June 26, 1995.
Mikhail Markelov, “Shamil Basayev: ‘I Am Not Planning Another
Budyonnovsk’,” Russian Press Digest (Novaya Gazeta), September 15, 1995.
Emir Dokka Abu Usman: “This Year Will Be Our Offensive Year,” Kavkaz
Center, May 17, 2009. Available from http://kavkazcenter.com/eng/content/
2009/05/17/10700 print.html (accessed August 15, 2012). Umarov had
expressed a similar sentiment already in 2005, after his own family was targeted.
See “We’re Beginning the War on the Territory of Russia,” Kavkaz Center,
56
Kristin M. Bakke
Thus it looks like radical tactics aimed at civilians have become not
just more common over time, but possibly more accepted on the part
of the fighters. Indeed, both the Dubrovka/Nord-Ost theater siege and
the Beslan school hostage crisis appear to have been more planned
as attacks against civilians than the 1995–1996 large-scale attacks.
Such a change in use and acceptance of radical tactics would coincide
with the entry of transnational insurgents, but it is important to note that
in both quotes above, the actions of the Russians serve as a justification.
The second war was characterized by a brutal campaign, including
civilian targeting, on the part of the Russian forces (e.g. Politkovskaya
2001, 2003); we thus cannot ignore the possibility that the turn to more
radical tactics on the part of the Chechens was a purely domestic response
(Speckhard and Ahkmedova 2006). We should also not overlook
the possibility that the growing kidnapping-for-ransom industry in
Chechnya, beginning in the interwar years, was driven purely by the
pursuit of profits by criminals (Murphy 2004, 139; Tishkov 2004,
107–126; Zürcher 2007, 105), rather than the influence of outsiders.
Yet if we accept that transnational insurgents over time contributed
to the Chechen insurgents’ turn towards increasingly radical tactics,
what is the evidence for diffusion mechanisms? The hostage crisis that
most prominently featured the influence of radical Islam was the
Dubrovka/Nord-Ost theater siege in October 2002, where the
hostage-takers ahead of time had made a video where they proclaimed
they were seeking martyrdom in the name of Allah. The video, which
featured women covered in black veils with Arabic script, aired on Al
Jazeera during the attack. Timing-wise, it is plausible that the tactics
were a result of learning or emulation via both relational and mediated
diffusion from transnational insurgents, as both training camps and
Salafi schools had been set up in Chechnya in the mid to late-1990s –
although there is little information about the kinds of tactics advocated
by the transnational insurgents in these fora. According to one account,
Khattab put videos of suicide bombings against Russian military
barracks online and trained students in hostage-taking techniques in
his camps (Murphy 2004, 33, 39), suggesting these were tactics he
May 9, 2005. Available from www.kavkazcenter.com/eng/content/2005/05/09/
3778.shtml (accessed August 15, 2012). See also this 2003 interview with another
Islamist Chechen commander, Amir Ramzan: “Next Year the War Will Seize the
Entire Caucasus,” Kavkaz Center, November 28, 2003. Available from www.
kavkazcenter.com/eng/content/2003/11/28/2039.shtml (accessed June 7, 2011).
Copying and learning from outsiders?
57
advocated. Similarly, while not independently confirmed, an investigation under President Maskhadov implied that Khattab was behind
the kidnapping and killing of six international Red Cross workers
in December 1996,29 again suggesting an acceptance for targeting
civilians. Khattab’s successor, Abu Walid al-Ghamdi, who had been
in Chechnya since the late 1990s but took over Khattab’s role when
Khattab was killed in the spring of 2002, is reported to have emphasized
suicide attacks in Russia over guerilla warfare in Chechnya as an
effective tactic (Vidino 2005). In 2003, the second-in-command of
the jihad in Chechnya issued a fatwa sanctioning the use of female
suicide bombers (Henkin 2006, 198). Moreover, according to Georgian
officials, by early 2002 a number of Arab fighters had fled from
Afghanistan to the Pankisi gorge in Georgia, located on Chechnya’s
doorstep, where several hundred Chechen and other insurgents were
trained in the use of toxic gases (Vidino 2005; Civil Georgia 2003).30
By the time of the Dubrovka theater siege in the fall of 2002, then,
these reports suggest that the Chechens may have been exposed to and
trained by transnational insurgents advocating radical tactics through
both relational and mediated diffusions.
Others have suggested that non-relational diffusion also played a
role, through news accounts about suicide terrorism in Iraq (Reuter
2004, 6). Indeed, in this respect, Khattab’s training camps served as a
forum for non-relational diffusion, as the students reported to have
seen videos of fighting in Palestine and Kashmir (Sokirianskaia 2010,
213); at least in Palestine, suicide terror had been a tactic since 1994.
As for the domestic insurgents’ motivation to adopt or adapt to new
tactics, the purpose of the Chechens’ video from the Dubrovka theater
siege (and a similar video associated with the Beslan school hostage
crisis), observers have argued, was to attract funding from external
sources in the Middle East (Wilhelmsen 2004, 45; Speckhard and
Ahkmedova 2006, 11). Thus, to the degree that diffusion took place,
it may have been emulation driven by strategic funding concerns,
rather than an internalized learning process.
29
30
“Warlord Khattab Implicated in Murder of Six Red Cross Workers in 1996,”
BBC Summary of World Broadcasts (ITAR TASS), March 27, 2000.
Basayev claimed that the so called militants in the gorge were actually Chechen
refugees. See “Chechen Field Commander Marries Third Wife, Gets Russian
POWs as Wedding Gift,” BBC Monitoring Trans Caucasus Unit (Kavkaz
Tsentr), December 14, 2000.
58
Kristin M. Bakke
Just as there is a domestic alternative explanation for shift in
framing, there is also a local template for hostage-taking as a tactic.
Indeed, hostage-taking has a long tradition in Chechnya, going back
to the North Caucasian people’s resistance to Russian annexation
in the 1700s and 1800s, where both the local population and
the Russians resorted to such tactics (King 2008, 53–59). Suicide
terrorism, in contrast, does not have a local historical template
among the Chechens, despite centuries of conflict with central rulers.
Thus in the absence of outside influence, it is unlikely that the
Chechens would have turned to such a tactic. Indeed, despite the
effect of domestic violence-begets-violence dynamics and the historical template of hostage-taking, the fact that some of the large-scale
hostage attacks in the Chechen wars have been aimed at attracting
funding from the Middle East suggest that transnational factors
have played a role in radicalization of tactics, but perhaps a smaller
role than what they are given credit for – especially by the Russian
government.
Effect on resource mobilization
Most accounts of transnational insurgents in Chechnya suggest
that their key contribution to the domestic resistance movement’s
resources is not manpower but access to financial resources, primarily
in the Middle East, although they have also brought along expertise in
communications and use of weapons (Wilhelmsen 2004, 41–46;
Vidino 2005; Ware 2005; Hughes 2007; Moore and Tumelty 2008).
Initially, the promise of the transnational insurgents who arrived
in February 1995 was their added resources, including both weapons
and access to finance, and the know-how and training expertise they
brought via relational diffusion in training camps and by fighting
side-by-side. In those early days of the war, the Chechen insurgents
found themselves overwhelmed against the Russians’ airpower.
Khattab and his followers were key participants in the Chechens’
retaking of Grozny in summer 1996, although the brutal winters in
the Caucasus mountains and the lack of knowledge of Russian
prevented large numbers of transnational insurgents from joining
and contributing to the Chechens’ struggle. Indeed, while Basayev
admitted to being assisted by Arab fighters in the June 1995 attack
on Budennovsk, when asked where he got his weapons, he did not
Copying and learning from outsiders?
59
mention foreign sources. Instead, Basayev emphasized that he was
buying his weapons from the Russians.31
In the interwar years, the ability of the transnational insurgents to bring
both competence and capacity increased. The Life and Times of Khattab
features Khattab saying that after the first war, he was asked by the
military and civilian Chechen leadership to help train insurgents, as the
Chechens doubted that the Russians would completely withdraw. Khattab and his crew established training camps in Serzhen’-Yurt, contributing to resource mobilization via relational diffusion. Because financial
resources in the immediate aftermath of the war were limited, Khattab
notes that they initially had to limit the number of people they were
training. However, as they made progress, they accepted up to 400 young
men per course, not only from Chechnya but also from Ingushetia,
Kabardino-Balkaria, Karachay-Cherkessia, Dagestan, Tatarstan,
Uzbekistan, and elsewhere. Sokirianskaia’s ethnographic study includes
interviews with Chechens who attended these camps:
After training in the first camp, the best were selected and transferred to the
military camp. Guys from Russia were taught mining, explosives, and the
like . . . This was real military training, these people knew that there would be
another war, they were preparing (quoted in Sokirianskaia 2010, 213).
Besides providing competence through his training camps, Khattab did,
per his own account, also provide humanitarian relief to the war-torn
population via his camps – which furthered his hero image among some
Chechens. He continued to fight alongside the Chechens until he was
assassinated in 2002.
The transnational insurgents’ contributions to the Chechen insurgents’
resource mobilization has also been via mediated diffusion, i.e. their link
to funding sources outside Chechnya, especially charities in Saudi Arabia
(e.g. Murphy 2004, 140–155; B. Williams 2005a). Throughout the first
war, Sheikh Fathi, who recruited Khattab to Chechnya, continued to
recruit transnational insurgents and was, reportedly, instrumental in
channeling funds from the Middle East to the Chechen insurgents. Khattab
himself, attuned to the power of propaganda, released tapes of the
Chechens’ struggle through a network of mosques, which helped recruitment of foreign fighters to Chechnya (Vidino 2005; Tumelty 2006).
31
“Basayev Takes Firm Stand on Chechnya’s Independence,” BBC Summary of
World Broadcasts (Segodnya), July 26, 1995.
60
Kristin M. Bakke
The Chechens used the same strategy in the second war, releasing tapes
from the Dubrovka/Nord-Ost theater siege in 2002 and the Beslan school
hostage crisis, portraying their struggle as an Islamist one. These tapes were
aimed at attracting funding from the Middle East (Wilhelmsen 2004, 45;
Speckhard and Ahkmedova 2006). Without having access to the resistance
movement’s budgets, it is hard to know where their funding actually comes
from, but Andrei Smirnov, a journalist covering the North Caucasus,
reports that a Chechen field commander in 2003 admitted to receiving
funding from international Muslim foundations (Smirnov 2007d).32 The
Chechen field commander Salman Raduyev stated in 1998 that his group
was funded by Islamic parties in the Middle East (Muzaev 1998b). More
generally, Khattab was an important broker establishing links between the
Chechen insurgent movement and Islamic charities (B. Williams 2007,
162). Similarly, the Dagestani Wahhabi preacher Kebedov, who entered
Chechnya in the interwar years, was allegedly funded by charities in Saudi
Arabia, Pakistan, and the US (Vatchagaev 2007).
The flow of resources may, in turn, have influenced tactical innovation,
especially the large-scale attacks such as the Dubrovka/Nord-Ost and
Beslan sieges. There are also accounts suggesting that the shift in framing,
towards a more religious message, has been affected by concerns for
resources. In other words, the domestic movement’s concerns for resources
may be a catalyst for shifts in the other two processes of mobilization.
That trend has not necessarily been welcomed by the domestic population.
The Chechen doctor Khassan Baiev notes in his autobiographical account
that: “We welcomed the humanitarian aid we received from Middle
Eastern countries, but we did not like it when they told us our Islam was
not the true Islam. For 400 years we have fought against people telling us
what to do” (2003, 206). Even Umarov has had to balance his message, so
that he can gain, or at least keep, support among Chechens who may be
motivated by self-determination and also attract funding from prospective
funders in the Middle East, especially Turkey, who may be skeptical of his
true commitment to an Islamist struggle (Smirnov 2007a).
The flow of resources via transnational insurgents has also influenced
the balance of power between domestic and transnational insurgents,
which in turn has affected both framing and tactical innovation in a
32
See the interview in “Next Year the War Will Seize the Entire Caucasus,”
Kavkaz Center, November 28, 2003. Available from www.kavkazcenter.com/
eng/content/2003/11/28/2039.shtml (accessed June 7, 2011).
Copying and learning from outsiders?
61
radical direction. In the first war, Dudayev was in control of most
funding sources, which came from within Russia. In the interwar years,
Maskhadov, more cash-starved in an already war-torn republic, became
dependent on warlords with external funding bases, thus giving both the
warlords and their foreign funders more power (Wilhelmsen 2004, 40,
46). Indeed, in fall 2007, the resistance leader in neighboring KabardinoBalkaria, Anzor Astemirov, posited that he was the one who had convinced Umarov to completely abandon the idea of a Chechen-centered
struggle and declare the Caucasus Emirate. His leverage was Umarov’s
dependence on non-Chechen fighters (Smirnov 2007c).
This attempt to trace resources suggest that the route through which
transnational insurgents influence the domestic insurgent movement
may start with the resources they bring to the table (or battlefield). If
shifts in framing and tactical innovation follow from a wish to attract
resources, the motivation causing shifts in these processes is about
strategic emulation rather than genuine learning.
Lessons learnt and further research
While the transnational dimensions of intrastate conflicts have received
a great deal of attention among both scholars and policy-makers in the
last few years, relatively little research has explored how transnational
actors influence such struggles. In this study, I theorize and trace
the diffusion mechanisms through which one group of such actors,
transnational insurgents, have influenced mobilization in the Chechen
separatist struggle against the Russian central government. I argue that
through both relational and mediated diffusion, which engender either
ideationally or instrumentally motivated learning and emulation,
transnational insurgents can affect a domestic movement’s framing of
goals, tactical innovation, and resource mobilization. While researching mechanisms in a civil war setting can be a challenging task due to
data limitations – it is, for instance, difficult to get reliable information
about an armed group’s training, much less whether the students
genuinely learnt something from that training – I have sought to
overcome these challenges by highlighting the observable implications
of my argument and triangulating data from a variety of sources (see
also the discussion in Checkel’s introductory chapter).
There are lessons in this study for both scholars and policy-makers. For
scholars, the study shows that mechanisms highlighted in relatively
62
Kristin M. Bakke
peaceful settings by the transnationalism and social movement literatures
also apply to more violent realms. For example, just like epistemic communities can influence the framing of a state’s foreign policy, transnational insurgents can affect the framing of a domestic movement’s
goals. Insurgents, too, copy or learn from others. While the emerging
scholarship on the transnational relations of civil wars so far has largely
assumed that transnational insurgents make such conflicts more likely or
more violent, this study draws on the social movement literature to
highlight that there are different aspects of a domestic movement’s
struggle that can be shaped by transnational insurgents – framing of
goals, tactical innovation, and resource mobilization. Yet my empirics
also point to conditioning variables and alternative explanations, emphasizing the need to carefully consider domestic factors that may either
shape or overshadow the role played by transnational insurgents.
For policy-makers, the study suggests that the role of transnational
insurgents should not be overstated. Rather than assuming that transnational insurgents influence domestic insurgents and insurgencies,
policies ought to be based on a careful examination of if and how
these actors influence the different processes in a domestic movement’s
mobilization effort. Transnational insurgents do not necessarily have
unidirectional and identical effects across these processes.
The next step in this research agenda picks up on these varied effects
and explores how outsider-induced changes in a domestic movement’s
mobilization processes are received among the local population.33 My
research so far suggests that it is not a given that transnational insurgents
actually strengthen the domestic movement, as the changes they encourage can cause a backlash. While this chapter focuses on how transnational insurgents may influence intrastate struggles by affecting
processes internal to the domestic movement, future research should
also explore how transnational insurgents, like other activist groups,
may shape intrastate conflicts by altering the external political context in
which the struggle takes place (see Meyer and Whittier 1994). Indeed, to
the degree that concerns about transnational terrorism changes states’
policies towards domestic challenges within their borders, these domestic challenges are more indirectly affected by transnational insurgents.
33
Kristin M. Bakke. “Acceptance and Resistance to Foreign Ideas: Transnational
Insurgents’ Impact on the Chechen Separatists.” Unpublished manuscript.
Department of Political Science, University College London. 2010.
|
3
Mechanisms of diaspora mobilization
and the transnationalization of
civil war*
fiona b. adamson
Introduction
As Checkel notes in his introduction to this volume, the roles of
external and transnational actors are attracting increased attention in
studies of civil war. Diasporas are unique in this regard as they consist
of populations and actors that are both external and internal. They
exist “outside the state . . . but inside the people” (Shain and Barth
2003, 451). As such, they epitomize the complexities of disentangling
processes and actors in the study of transnationalism and civil war.
The aim of this chapter is to contribute to a broader understanding
of the transnational dynamics of civil war through a study of diaspora
mobilization and conflict. The chapter does so by briefly pointing to
the existing literature, before focusing on several causal mechanisms
that can help us to disentangle and disaggregate aspects of diaspora
mobilization that may exacerbate violent conflict in the home state.
I then employ these mechanisms to examine the case of the conflict
between the Kurdistan Workers’ Party (PKK) and the Turkish state in
the 1980s and 1990s. I conclude by discussing both the contributions
and limits of focusing on mechanisms of diaspora mobilization in civil
war, placing them within a broader context that includes other factors
such as geopolitics and state interests.
Diasporas and violent conflict
In recent years, there has been increased interest in the role that
diaspora populations can play in contributing to or exacerbating
* Thanks to Kristin Bakke, Andrew Bennett, Jeffrey Checkel, Matthew Evangelista,
Scott Gates, Kristian Harpviken, Sarah Lischer, Elisabeth Wood, and the other
project participants for their many helpful comments on the chapter. Elena
Fiddian Qasmiyeh, Maria Koinova, Francesco Ragazzi, and Idean Salehyan also
provided a number of useful insights and suggestions on earlier drafts.
63
64
Fiona B. Adamson
violent conflicts.1 Research on this topic ranges from quantitative
studies that operationalize diasporas as independent variables (Collier
2000; Collier and Hoeffler 2004), to collections of individual case
studies (Smith and Stares 2007), to broader theoretical and analytic
discussions (Anderson 1998; Huntington 1996; Kaldor 1999). A loose
collection of studies is emerging that is pushing forward a research
agenda on diasporas as potential initiators or exacerbators of violent
conflict, as well as mediators of conflict and/or actors in conflict resolution and post-conflict reconstruction (Adamson 2004; 2005; 2006;
Brinkerhoff 2006; Koinova 2010; Shain and Aryasinha 2006).
The most cited finding of this research has been that of Collier and
Hoeffler (2004), who in a quantitative study analyzing “greed” vs.
“grievance,” identified diasporas as a significant force in fueling violent
conflict. Building on earlier observations (Angoustures and Pascal 1996)
regarding the financing of violent conflicts by diaspora groups, Collier
and Hoeffler attempted to measure quantitatively the impact of diaspora
populations on civil war by correlating the size of a country’s diaspora to
the incidence of civil war in that country. They suggested that the correlation could be explained by the ability of diasporas to provide financial
support to rebel organizations through donations, and argued that this
finding lent confirmation to their overall argument that “greed” or
political economy factors trumped “grievances” in explaining civil wars.
This quantitative finding that diaspora populations may influence
the course of violent conflict appears to be supported anecdotally
by numerous other studies. In discussions of how the dynamics of
conflicts are changing due to broader patterns of globalization in
the international system, Kaldor (1999) notes both the ideational and
material impacts of diasporas on conflicts around the world. Support
is also provided in a RAND survey of sources of external support in
contemporary insurgencies, which noted that one of the major changes in
the international security environment post-Cold War was the increased
importance of non-state actors in general, and diasporas in particular, as
sources of financial and material support for insurgencies and guerrilla
movements (Byman et al. 2001, 41). Others have portrayed diaspora
1
There is a broader literature addressing the question of what constitutes a
“diaspora” and how to define it (i.e. Adamson and Demetriou 2007, 2012;
Brubaker 2005; Cohen 1997; Dufoix 2008; Sheffer 2003). I do not directly
address these debates here, although the framework used in this chapter is
informed by the literature.
Mechanisms of diaspora mobilization
65
populations as engaged in a form of “civilizational rallying” who come
to the aid of their “ethnic kin” in times of violent conflict (Huntington
1996) or foster “long-distance nationalism” that is characterized by
“a serious politics that is radically unaccountable” (Anderson 1998, 74).
The implication here is that members of diasporas can involve
themselves as supporters of violent conflicts in their home countries,
without paying the consequences of living in societies marked by
political violence. These observations have recently been extended
more broadly to the phenomenon of terrorism – Sageman (2004), for
example, claims that 84 percent of those involved in al-Qaeda-inspired
terrorism have been recruited in a diasporic context, with the majority
of recruitment taking place in Western Europe. In addition, there is a
growing body of empirical studies of particular conflicts (Biswas 2004;
Danforth 1995; Fair 2005; Gunaratna 2001; Ho 2004; Hockenos
2003; Lyons 2006; Rapoport 2003; Shain 2002; Smith and Stares
2007) that have examined the extent to which members of diaspora
groups have been active supporters of political violence.
Yet, although a growing body of literature has established the
significance of diaspora populations for understanding the dynamics
of violent conflict, there has to date been little attention paid to
systematizing the study of such diaspora involvement. The conclusions
of Collier’s (2000) study, for example, relied on the finding that there is
a correlation between US Census data on foreign-born populations and
the recurrence of violence in civil wars. While it suggested reasons for
this finding it did not attempt to trace the various processes by which a
population abroad would impact on a conflict in its “home” country.
Similarly, many qualitative studies that have noted the importance
of diasporas have been largely descriptive and have not attempted
to identify causal mechanisms that could be used to systematically
compare diaspora involvement in conflicts across different cases.
Part of the problem in investigating the processes and causal mechanisms by which diasporas come to play a role in conflicts may have
to do with how they are conceptualized in much of the literature – as
unitary actors treated as “independent variables.” Without examining
intervening processes of political mobilization, it is difficult to arrive
at conclusive statements regarding the role that diasporas play in the
international security environment. Indeed, some research (i.e. King
and Melvin 2000) convincingly shows that in many cases diaspora
mobilization should be treated as a dependent variable rather than an
66
Fiona B. Adamson
independent variable, i.e. that diaspora mobilization is a consequence
rather than a cause of violent conflict.
A focus on political mobilization points to parallels between the
processes of transnational mobilization that are seen in diaspora politics
and other instances of transnational politics (Adamson 2002; 2005;
2012). As Checkel notes in his introduction, there is now a wellestablished literature in international relations, comparative politics and
sociology that theorizes transnational processes and movements (Bob
2005; Della Porta and Tarrow 2004; Keck and Sikkink 1998; Khagram,
Riker, and Sikkink 2002; Tarrow 2005). Similarly, there is growing
literature on civil wars and ethnic conflict that has begun to examine their
transnational dimensions (Wood, this volume; see also Gleditsch 2007;
Gledistch and Salehyan 2006; Hironaka 2005; Jenne 2007; Salehyan
2009). What is needed is to combine insights from such literatures in ways
that can better specify the processes and causal mechanisms by which
diaspora politics come to impact on the dynamics of violent conflicts.
Paying close attention to causal mechanisms is important, because one
need only examine the wider body of research in diaspora studies to quickly
see that the purported effects and outcomes that are attributed to diasporas
vary widely. Diasporas, for example, can affect policy-making outcomes in
their “host state,” particularly in the area of foreign policy-making (Weil
1974; Mathias 1981; Tucker et al. 1990; Uslaner 1991; Huntington 1997;
Smith 2001; Saideman 2001). They can enhance economic development or
contribute to democratization and the observance of human rights (Koinova 2009; Van Hear 1998; 2002; Shain 1999). Such a wide range of
possible impacts would suggest that diasporas in and of themselves do not
inherently contribute to any particular outcome – their activities and
political impacts vary. This calls into question the utility of conceptualizing
diasporas as unitary actors that have fixed effects on outcomes. A more
fruitful approach would be to assume that the activities of members
of diaspora populations vary, and then to ask under what conditions and
for what reasons they are most likely to become participants in civil wars.
This involves identifying a set of key causal mechanisms.
Diaspora mobilization in civil wars: identifying
causal mechanisms
A focus on causal mechanisms and the use of process tracing in case
studies provides an important starting point in helping us to specify
Mechanisms of diaspora mobilization
67
exactly when, why, and how diaspora populations may play a causal role
in the initiation or perpetuation of civil wars. This involves moving away
from the use of primordial and essentialist assumptions that reify
“diasporas” as unitary actors, and instead examining the actual processes
by, and conditions under which, members of diaspora populations
become mobilized as de facto participants in a “long-distance conflict.”
Much of the literature on political mobilization, social movements,
and contentious politics has focused on processes that occur at
the domestic level or within states. Similar mechanisms are found in
processes of political mobilization that occur in a transnational setting
and that take place across state borders (Della Porta and Tarrow
2004; Keck and Sikkink 1998; Khagram, Riker, and Sikkink 2002;
Tarrow 1998; 2005; see also Bakke, this volume). However, as Bennett
reminds us, a focus on mechanisms does not mean that we should
ignore the broader structural context, such as global inequalities and
variations in domestic political opportunity structures (Bennett, this
volume). Rather it allows us, within a given set of broader structural
conditions, to explicate the specific dynamics and processes that link
dispersed populations and that mobilize them to collectively engage in
particular sets of actions that affect the course of a violent conflict.
As an initial step, one can separate the study of diaspora involvement
in violent conflicts into two stages: the process of diaspora mobilization,
and the impact of diaspora mobilization on the course of a conflict. In the
first stage, the mobilized diaspora is treated, loosely, as the dependent
variable. In other words, the phenomenon to be explained is how and
why transnationally dispersed populations come to be mobilized to be
engaged in a violent conflict thousands of miles away. In the second
stage, the mobilized diaspora becomes the independent variable. The
focus is on how mobilized diaspora populations then come to impact on
the onset, duration and intensity of a violent conflict. By separating
diaspora involvement in violent conflict into its constituent parts, we
gain greater analytical leverage and can begin the process of exploring
variations in each of the mechanisms in greater detail.
In the following sections, I identify five causal mechanisms that
provide a starting point for understanding how and when diaspora
populations become actors in civil wars. Like Checkel and Gerring,
I define a causal mechanism as “a pathway or process by which an
effect is produced or a purpose is accomplished” (Checkel, this
volume). In actuality, of course, there are numerous and multiple
68
Fiona B. Adamson
mechanisms that could play a role in diaspora mobilization and
the transnationalization of conflict; many have been identified and
elaborated upon by scholars working on transnationalism and/or contentious politics (McAdam, Tarrow, and Tilly 2001; Tarrow 2005).
Other factors, such as the origins and process of migration, state
repression, and human rights abuses in the country of origin, and the
emergence of political opposition in both the home state and the
diaspora, are important for understanding diaspora involvement in
violent conflict, but will not be the primary focus of this chapter.
The five causal mechanisms examined here are divided into two
subsets. The first set – transnational brokerage, strategic framing, and
ethnic or sectarian outbidding – are related to diaspora mobilization.
The second – resource mobilization and lobbying and persuasion – relate
to the impact of the mobilized diaspora on the conflict. It may be tempting
to treat the two elements of mobilization and impact as two stages in a
sequence (first the diaspora is mobilized, and then the mobilized diaspora
impacts on the conflict), however, in reality the two processes may
overlap, reinforce each other, or create feedback loops in ways that do
not lend themselves to a strictly sequential mapping or analysis.
The process of diaspora mobilization
It is common in quantitative approaches to civil war and ethnic conflict
to reify a particular collective actor – such as an ethnic group or
organization – and to treat the group as “acting” in a conflict. However,
beginning with the assumption of a pre-existing collective actor begs the
prior question of how that collectivity itself becomes constituted and
comes to be viewed as an “actor” and why that collective actor comes to
have an “interest” in a conflict that is occurring primarily in a different
geographical locale. In other words, it is necessary to look at questions
concerning the formation of collective agency, group identity, and
collective interests/preferences as part and parcel of our understanding
of conflict processes. A mechanism-based approach can help us shed
light on the micro-dynamics of this process.
Transnational brokerage
Social network theorists have identified “brokers” as particularly
powerful actors in their ability to link together disparate networks.
Mechanisms of diaspora mobilization
69
When two networks are separated by a “structural hole” a broker can
gain power by filling the gap and bringing together two unlinked
networks (Burt 2005; Goddard 2007; Padgett and Ansell 1993).
A focus on the mechanism of transnational brokerage helps explain
how networks defined by diasporic ties become connected with
“conflict networks” that are actively engaged in political violence.
A necessary yet insufficient condition of diaspora involvement in conflict is linking of diasporic networks in one or more so-called “host
states” with those networks that perpetuate or sustain conflict in the
“home state.”
A broker plays a role in connecting a group or network symbolically
but also materially to a conflict. Here one can compare with the
literature in social movements that examines how transnational
brokerage brings about a “scale shift” in incidents of transnational
contention (McAdam and Tarrow 2005; Tarrow 1998; 2005). To
apply this to civil wars, it is necessary to understand how transnational
brokers emerge and connect diasporic networks in different states and
to conflict networks in the so-called “home state.” This is a prior step
to understanding how ideas, material resources, expertise, recruits, or
other elements are transferred back and forth between the diaspora
and the conflict zone.
Strategic framing
The concept of “strategic framing” has been largely developed in and
drawn from the literature on social movements (Benford and Snow
2000). However, it is also found separately in work on policy frames
(Rein and Schoen 1996), as well as being increasingly employed as a
useful analytical tool by scholars of international relations (Bob 2005;
Busby 2007; Khagram, Riker, and Sikkink 2002; Payne 2001); it also
plays a central role in the chapters by Bakke and Schmitz in this
volume. Frames have been defined as “schemata of interpretation”
that are “intended to mobilize potential adherents and constituents,
to garner bystander support, and to demobilize antagonists” (Benford
and Snow 2000, 614; Goffman 1974, 21, as cited in Benford and
Snow 2000) or as persuasive devices used to “fix meanings, organize
experience, alert others that their interests and possibly their identities
are at stake, and propose solutions to ongoing problems” (Barnett
1999, 25, as cited in Payne 2001, 39). Social movement approaches
70
Fiona B. Adamson
conceptualize framing as “the conscious strategic efforts by groups
of people to fashion shared understandings of the world and of
themselves that legitimate and motivate collective action,” and have identified the linking of different frames, or “frame alignment” as an important
way to explain the emergence and growth of social movements (McAdam,
McCarthy, and Zald 1996, 6 as cited in Busby 2007, 251).
The process of diaspora mobilization is dependent on the ability of
actors to devise and articulate frames that resonate with members of
diaspora groups in ways that successfully align the perceptions, values,
or interests of diaspora members with those of civil war actors or
conflict entrepreneurs. Deploying notions of national belonging and
duty or “kinship” could be one possible framing strategy, but there are
certainly other powerful frames that could be deployed to mobilize
members of a diaspora. For example, some studies have noted that
diaspora entrepreneurs can successfully deploy frames emphasizing
guilt or obligation in order to tie members of diasporas to their country
of origin, such as the obligation to send remittances to family
members.2 A focus on strategic framing processes emphasizes that
involvement by individuals in so-called “homeland politics” or a conflict
thousands of miles away is a puzzle to be explained rather than a
situation to be expected.
Ethnic or sectarian outbidding
Ethnic outbidding broadly refers to the politicization of ethnic differences by elites or political parties (Rabushka and Shepsle 1972;
Horowitz 1985). In dynamics of ethnic outbidding, parties or elites
attempt to outdo each other, leading to a cycle of polarization that
fuels extremism (Brubaker 1996; Gagnon 1994; Chandra 2005).
Outbidding can involve the use of violence, both as a means of
demonstrating strength and resolve, and as a way of marginalizing
more moderate alternative groups. The taking of extreme positions
can be seen as a means of increasing one’s bargaining leverage (Jenne
2007). The aim of the process of outbidding, including the use of
repression and violence, is to draw a firm distinction between what
2
Laura Hammond. “Obliged to Give: Remittances and the Maintenance of
Transnational Networks Between Somalis ‘At Home’ and Abroad.” Unpublished
Manuscript. 2007.
Mechanisms of diaspora mobilization
71
Kydd and Walter (2006) have referred to as “zealots” vs. “sellouts.”
In other words, the taking of extreme positions – either rhetorically
or through the selective employment of violence – can, under certain
conditions, mobilize public support, thus leading to increased radicalization. It is this strategic deployment of both language and action
to simultaneously increase power and legitimacy (thus creating a feedback
loop in the form of additional support – material as well as social) that
makes this mechanism interesting.
Diaspora organizations can operate in a manner similar to ethnically
defined political parties (sometimes they are directly affiliated with
home-state political parties). Indeed, they may have increased incentives
to do so to retain support. Because the “diaspora” as a construct is
defined according to an ethnic or national category (or, at the minimum,
a connection to a real or imagined homeland), this may create incentives
for diaspora organizations to attempt to outbid each other in their
articulation of a national or ethnic identity as a means of increasing
their power and standing in the diaspora. When organizations are
competing for diaspora support, incentives may exist for ethnic outbidding. This is arguably even more likely to occur if there is instability,
political uncertainty or conflict in the so-called home state. After all, in
some respects the “diaspora” ceases to exist if a distinct identity is not
mobilized and fostered, as arguably occurred with the “Irish diaspora”
in the US following the 1998 Good Friday Agreement (Cochrane 2012).
The impacts of mobilization on violent conflict
Once a process of mobilization takes place, how does a transnationally
mobilized population impact on violent conflict? Two additional
mechanisms – resource mobilization and lobbying-persuasion – help
to explain the direct and indirect impacts a mobilized diaspora can
have on a conflict. These mechanisms are drawn directly from the
literatures on social movements and transnational advocacy
networks, and are common to both non-violent and violent forms of
mobilization (McCarthy and Zald 1977; Pichardo 1988).
Resource mobilization
Diaspora mobilization emerges as a strategy in civil war when
members of diaspora populations have access to needed resources.
72
Fiona B. Adamson
Political entrepreneurs can engage in resource mobilization activities
and then channel them into the support of conflict-related activities.
Resources can include capital and financial resources, but also human
resources – technical expertise, or even recruits. Identifying the mechanism of resource mobilization provides a means for examining the
processes and pathways by which diaspora organizations manage to
mobilize resources and channel them both to diasporic activities and
directly to the zones of conflict.3
Lobbying and persuasion
Lobbying and persuasion are additional mechanisms by which actors
in the diaspora can impact the course of violent conflict. There are
numerous studies of persuasion or argumentation in the field of international relations (Checkel 2007; Johnston 2008; Price 1998; Risse
2000). One of the most applicable models to diaspora involvement in
civil wars is that of the “boomerang pattern” (Keck and Sikkink 1998,
12–13). In this model of transnational lobbying, NGOs in a state
where political mobilization is difficult will network with NGOs outside it. In turn, these external NGOs place pressure on their own
governments and on international organizations, which then place
pressure on the original state. This is a useful model for understanding
how actors in a civil war situation can turn to diasporic elites to lobby
internationally on their behalf (see also Hamberg, this volume).
Diaspora mobilization in the Kurdish conflict in Turkey
(1980–2000)
Having identified a set of causal mechanisms, I now turn to their
application in a case study of the Kurdish conflict in Turkey, which
has been notable for its high level of diaspora involvement. The civil
war between the Kurdistan Workers’ Party (PKK) and the Turkish
state has claimed approximately 40,000 lives since 1984, has resulted
in thousands of villages being destroyed, and has produced an estimated
3 million refugees/displaced people. It is a conflict for which there is
extensive data, including a number of academic reports, as well as
3
For a similar understanding of the role played by resource mobilization in other
instances of transnationalized civil war, see Bakke and Schmitz, both this volume.
Mechanisms of diaspora mobilization
73
studies of the dynamics of Turkish and Kurdish transnational politics
(e.g. Barkey and Fuller 1998; Lyon and Ucarer 2001; stergaardNielsen 2001; 2003; Romano 2006; Wahlback 1999; Watts 2004;
White 2001).
By utilizing these existing studies, and supplementing them with
primary research such as the use of press reports, government documents, and ethnographic research, it is possible to engage in a processtracing exercise that sheds light on both the mechanisms of diaspora
mobilization and the impacts of this mobilization on the conflict. This
case study does not present a full overview of the conflict, its various
stages, iterations and complexities, and relation to other developments
in Europe and Turkey. Rather, my aim is to show how the identified
causal mechanisms shed important light on the transnational dimensions
of the Kurdish conflict in Turkey during the period of 1980–2000.
The historical roots of the contemporary conflict can be traced to the
founding of the Turkish Republic with the Lausanne Treaty of 1923.
As part of Turkish nation-building, expressions of Kurdish or other
minority identifications were severely repressed by the state and viewed
as threatening to Turkish national unity. An ideology of secular Turkish
nationalism, derived from the principles of Kemalism put forth
by Turkey’s founder, Kemal Atatürk, came to dominate the political
discourse in Turkey. In this context, there were several regional Kurdish
rebellions and uprisings within Turkey in the 1920s and 1930s, and low
levels of political activity around the Kurdish issue in Eastern Turkey
in the 1960s, but the most significant event was the founding of the
PKK by Abdullah Öcalan in 1978.4
The PKK began to agitate among Kurdish areas of southeastern
Turkey in the late 1970s. During these years, as Turkey descended
into a state of internal anarchy characterized by violent clashes
between the extreme left and right, the PKK began to attack targets
and established a presence in Kurdish areas of the country. The 1960s
and 1970s had seen a significant migration from Turkey to Western
Europe, especially Germany. This was a response to the labor shortage
in Germany following World War II. Between 1961 and 1974, the
number of Turkish workers in Germany grew from 2,400 to 617,500
4
For general background on the Kurdish conflict in Turkey, see McDowall 1996,
395 458; Barkey and Fuller 1998; Gunter 1990; Izady 1992, 67 72, 215 218;
Entessar 1992, 81 111.
74
Fiona B. Adamson
(Akgündüz 1993, 161); by 1974, their number was greater than that of
any other nationality. The number of Turkish citizens in Germany
continued to grow even after recruitment stopped in the 1970s, due
to a combination of family reunification policies and political asylum,
eventually reaching approximately 2 million.
In 1980, Turkey experienced a military coup and the PKK, along
with other left-wing organizations as well as trade unions, was severely
repressed, with the government banning a number of political parties
and arresting political activists. In 1982, the new Turkish constitution
strictly prohibited the use of the Kurdish language, Kurdish publications, and the establishment of Kurdish political parties or other manifestations of Kurdish identity. Following the 1980 military coup and
the severe repression that followed, the PKK regrouped in Syria and
Lebanon and formed an armed organization that mounted attacks in
southeastern Turkey beginning in 1984.
One of the key features of the Kurdish conflict in Turkey has been its
transnational dimensions and the active role that Kurdish populations
in Europe have played in it. This role has not remained static, but
instead has evolved over time. A full study of the transnational dimensions of the Kurdish conflict would have to account for the fluctuations
that occurred as populations from Turkey became more established in
Europe; as external events (such as the arrest of the PKK leader
Abdullah Öcalan in 1999) shifted the dynamics of the conflict; and
as the relationship between Turkey and Europe has evolved in the light
of European Union (EU) negotiations. It would also need to take note
of the changing political situation in Turkey with the introduction of
constitutional reforms due to the EU harmonization process, the easing
of restrictions on the Kurdish population, and the emergence since
2001 of the Islamic Justice and Development (AK) party as a major
political force. Nevertheless, it is possible to trace the basic contours of
diaspora involvement in the conflict by employing the five mechanisms
discussed above.
Transnational brokerage: linking Europe with the conflict
in southeastern Turkey
To understand the transnational dimensions of the civil war, it is
important to understand how populations in Europe became linked
with the conflict in Turkey. A key factor in this respect was the
Mechanisms of diaspora mobilization
75
emergence of a group of “transnational brokers” who managed to
connect migrant populations in Germany and elsewhere with conflict
networks and actors in southeastern Turkey. These brokers were
largely members of the National Liberation Front of Kurdistan
(ERNK), which was the political wing of the PKK, and which embedded
itself in populations in Europe and undertook mobilization activities in
support of the war in Turkey. A key development in this respect was the
exodus of a group of Kurdish intellectuals, activists, and militants to
Europe from Turkey following the 1980 military coup; this group
interacted with already-settled populations from Turkey in Europe.
Some of these activists began to organize a Kurdish nationalist movement in Western Europe and establish European branches of the ERNK
in a number of European states.
The vast majority of Kurdish cultural associations, information
centers, and publications during the 1980s and much of the 1990s were
associated with or controlled by European branches of the PKK through
the ERNK or its affiliates. Throughout the mid 1980s to early 1990s,
the political wing of the PKK operated legally in most of Europe, with
its above-ground cultural, social, and political organizations existing
side-by-side with a parallel covert and tightly organized underground
structure (Stein 1994, 86). The European Central Committee of the
ERNK had headquarters in Cologne and Brussels with national organizations in Germany, Belgium, France, Holland, England, Switzerland,
Italy, and the Scandinavian countries (Stein 1994, 91; Barkey and Fuller
1998, 38; Helfer 1988, 20).
Throughout the 1980s and 1990s, the PKK pursued a dual strategy.
It used political violence against the Turkish state (and segments of the
Kurdish population) by perpetuating an armed conflict in southeastern
Turkey, in which institutional voids in Lebanon and Syria and,
following the Gulf War, Northern Iraq, were used as bases to train
militants and plan incursions into southeastern Turkey. At the same
time, it engaged in political activities in Europe as a means of mobilizing populations that could provide resources and political support for
the conflict. Most of the Kurdish leadership in Germany during this
period was already connected in some way to the Kurdish student
activist networks that had arisen in Turkey in the 1970s. There was
thus an established link between Kurdish activists in Germany and the
PKK leadership that operated out of Syria and southeastern Turkey
(Barkey and Fuller 1998, 32). Members of this group of exiles in
76
Fiona B. Adamson
Europe maintained contacts with networks of PKK members who had
fled to Syria and Lebanon, and who had begun to take up arms against
the Turkish state from across the border. Exiles in Germany and other
countries in Europe managed to build a pan-European counterpart to
the PKK’s political wing, the ERNK, and successfully created a transnational structure that was organized as a network of local cells that
spanned the continent and beyond (Bundesamt fuer Verfassungsschutz
1996, 7). This Kurdish movement “became an almost invisible network
spread across the globe” (Bruinessen 1998, 48).
Thus, the political organization of the Kurdish diaspora was undertaken by political entrepreneurs who were well connected with the PKK
structures in Syria and southeastern Turkey – often the same individuals
were moving back and forth between Europe and Damascus. This
provided a transnational brokerage mechanism for directly connecting
diaspora populations in Europe with the PKK conflict networks in Syria
and southeastern Turkey. Although numerous social, familial, political,
and official ties connected Kurdish populations in Europe with Turkey,
the process of political mobilization around the conflict cannot be
understood without an account of the ERNK/PKK’s transnational
activities. As an organization, it successfully managed to cultivate transnational brokers who linked populations and resources in Europe to
the armed insurgency in southeastern Turkey and surrounding regions.
Strategic framing: constructing a politicized Kurdish identity
in Europe
The political activities undertaken by the transnationally connected
Kurdish activists in Europe involved mobilizing migrant populations
from Turkey as Kurds. To understand how this occurred, one has to
examine the use of strategic framing in the construction of a politicized
Kurdish identity amongst populations in Europe. In his provocatively
titled article, “How Turks Became Kurds, Not Germans,” sociologist
Claus Leggewie made the claim that “there have always been Kurds
among the Turkish ‘guest workers’ and refugees, but most of them
did not discover their ‘Kurdishness’ until they came to Europe”
(Leggewie 1996, 79). Similarly, Bruinessen argues, “among the Turkish
‘Gastarbeiter’ who had migrated to Western Europe since the late
1950s, there were many who ‘discovered,’ in the course of the past few
decades, that they were not Turks but Kurds” (Bruinessen 1998, 44–45).
Mechanisms of diaspora mobilization
77
Many who grew to identify themselves as Kurds in Europe may have
migrated to Germany or other countries as self-identified Turks,
grew up speaking Turkish, not Kurdish, or had parents who came
to Germany as labor migrants in the 1960s and viewed themselves
as secular and Kemalist Turks. This replicates some aspects of the
situation in Turkey, where “assimilated Kurds” often readily identify
themselves as Turkish, and even highly politicized Kurds may speak
Turkish, rather than Kurdish. In both Turkey and within the immigrant community in Western Europe, there have also always been
Kurdish-speakers and self-identified Kurds, but very few members of
this category drew upon their Kurdishness as a basis for political
activism during the 1960s or early 1970s.
The politicization of a Kurdish identity and the development of
a Kurdish nationalist movement within the “Turkish” immigrant community in Germany occurred largely during the 1980s and the 1990s.
This was due to the efforts of the transnational brokers who emigrated
from Turkey to Germany and other European countries following the
1980 military coup. When Kurdish political entrepreneurs arrived in
Germany and other European states in the late 1970s and early 1980s,
they had access to immigrant populations that provided a large pool of
potentially active Kurds. Early migration patterns from Turkey to
Germany produced networks that were disproportionately drawn
from the ethnically Kurdish regions of Turkey, due to the low
levels of economic development in southeastern Anatolia. By the mid
1970s and the ending of the official recruitment of workers, there
were significant portions of the “Turkish Gastarbeiter” population in
Germany that originated from ethnically Kurdish regions, yet most of
this population did not identify with a Kurdish political identity or
project. Bruinessen writes: “Until the late 1970s, relatively few migrant
workers emphasized their Kurdish identity. Most of the workers,
especially those of rural origins, were reluctant to become involved
with politics. Moreover, their European surroundings defined them
as Turks, and this remained in most contexts the relevant identity”
(Bruinessen 1998, 45).
When activists from the PKK arrived in Western Europe in the 1980s,
they went about explicitly constructing and politicizing a distinct
Kurdish national identity. Through processes of strategic framing, they
were able to create and deploy a new form of nationalism that appealed
to a population that felt doubly marginalized by the Turkish state and
78
Fiona B. Adamson
its official national narrative, as well as by European states. This was
particularly strong in Germany, in which second-and even thirdgeneration migrants were viewed as “foreigners” (Auslaender) who were
denied German citizenship, and often remained socially and economically
marginalized. At the same time, official agreements between Germany and
Turkey meant that this population was still living in the grip of the Turkish
state within Germany – whether via schooling provided by Turkish
teachers, religious life dominated by imams from the Turkish Presidency
of Religious Affairs (Diyanet), or Turkish officialdom in everyday life in
Germany, such as in the registration of births.
This encouragement of diaspora members to discover their Kurdish
identity and to construct Kurdishness as a form of resistance to state
hegemony in both Europe and Turkey had a certain appeal, especially
amongst segments of the second-generation Turkish immigrant population. A variety of frames and techniques were used to mobilize a
Kurdish national movement in Europe. First and foremost was the
attempt to elevate Kurdish language and culture to a status on par
with its Turkish counterparts. This was symbolically important as
an act of protest because of the restrictions on Kurdish linguistic
and cultural expression in Turkey at the time, but also because of
the continuing involvement of the Turkish state with “its” diaspora
in Germany. Absent jus soli/place-of-birth citizenship criteria in
Germany, births to Turkish citizens had to be registered with Turkish
authorities – meaning that any restrictions on giving Kurdish names
to infants in Turkey were replicated for Turkish citizens in Germany.
The promotion of Kurdish cultural activities, such as Kurdishspecific festivals and ceremonies became another form of symbolic
protest. Most important amongst these was the celebration of the
festival “Newroz” (New Year) which during this period was banned
in Turkey. Similarly, the publication of Kurdish-language books,
poetry, and newspapers, as well as the promotion of Kurdish music
and the establishment of a Kurdish-language satellite television that
could compete with Turkish stations were cultural activities that could
be freely pursued in Europe, but were also deeply political acts owing
to their illegality in Turkey (Hassanpour 1998). This promotion of
Kurdish culture, language, and customs became a political act in itself –
a form of resistance that carved out a publicly available alternative
identity category for members of the diaspora that posed a challenge
to the hegemonic categories of both “German” and “Turkish.”
Mechanisms of diaspora mobilization
79
The promotion of a Kurdish national identity, combined with the
radical politics of the PKK and its Marxist ideology that endorsed
armed struggle, fostered a culture of resistance in Europe that equated
Kurdish nationalism with resistance to Turkish and German
nationalism, American hegemony, capitalism, and traditional authority
structures (feudal and tribal systems). This frame of resistance to
existing power structures resonated with a large number of young
Turks/Kurds in Europe and, while it was not the only set of frames
available, it was particularly successful in appealing to a segment of
the population that was predisposed to identify culturally as Kurdish.
Kurdish political entrepreneurs in the 1990s were thus able to
formulate political categories and ideologies that presented individuals
with identity options that countered both German and Turkish national
identities. They constructed categories of belonging that resonated with
the lived experiences of many young members of the Turkish-Kurdish
community in Europe, and used this as a means for expanding their
base of political support. As Eccarius-Kelly (1999, 16–17) noted:
Once the PKK succeeded in establishing an emotional link with such a
subculture group [of German Kurds] it [was] easier to mold members into
PKK structures . . . In exchange for membership and activism within the PKK
structures, young Kurds are offered a Kurdish dream world complete with
myth of origin, sense of ethnic harmony, social status and respect.
The conflict itself helped to foster and reify a Kurdish identity – thus
creating a feedback loop between mobilization in the diaspora, conflict
activities in southeastern Turkey, and further mobilization. The fact
that Germany vocally denounced the PKK only added to the attraction
it had for many who joined “the movement to become part of an ‘elite
organization’ with an almost mythic fame for its defiance of the
German state” (Eccarius-Kelly 1999, 24).
Ethnic outbidding: the PKK and its alternatives
The PKK was not the only Kurdish group active in Europe in the 1990s
and it also faced stiff competition from a range of non-Kurdish entities
including Turkish nationalist, Turkish leftist, and religious organizations. A number of Kurdish organizations with links to Kurdish Social
Democrat movements in Turkey – who eschewed political violence –
also existed in Germany; they operated largely under the guise of the
80
Fiona B. Adamson
umbrella organization KOMKAR, which viewed itself as working
simultaneously for the rights of Kurdish immigrants within Germany
and for an improvement of the political situation of Kurds within
Turkey. In addition, there were some smaller Kurdish organizations
operating in Germany at the time. Generally, however, the PKK was
successful in acquiring a relative monopoly over the segment of the
“Turkish” immigrant community that came to identify as “Kurdish.”
This was accomplished through a combination of successful mobilization and propaganda, the creation or co-optation of an extensive
organizational structure, as well as coercion.
The relationship between the PKK and other Kurdish groups in the
1990s can be viewed as a case of the PKK winning a competition of
“ethnic outbidding” within Kurdish diaspora communities in Europe.
Under conditions of intense suppression of Kurdish identity in Turkey
and discrimination in Germany and Europe, the PKK was able to
devise a strategy that equated “Kurdish” with resistance, and not
assimilation. Thus, other groups were constructed as less “Kurdish”
to the extent that they were constructed as promoting integration into
German society or collaboration with the Turkish state. By taking an
extreme position, that included the use of contentious politics, resistance, and violence, the PKK established itself as the acknowledged
representative of the Kurdish cause in Europe. It was thus able to
gain hegemony over Kurdish politics in the diaspora, despite the fact
that many individual diasporans may have disagreed with its tactics
and many of its actions.
The PKK did not hesitate to use violence to suppress its rivals within
Europe. But while coercion did play some role in the PKK’s monopolization of the Kurdish movement, the high levels of support that it
had within Kurdish communities in Europe was also attributable to its
recognition as the only organization that could viably challenge the
interests of Turkish state elites. Some estimates put political support for
the PKK as high as 90 percent among those who identified as Kurds in
Europe in the 1990s, with at least 10,000 active members in the party
or related organizations and 50,000 active mobilized sympathizers in
Germany (Bundesamt fuer Verfassungsschutz 1996, 8; Barkey and
Fuller 1998, 32).
Thus, the PKK, as the most radicalized of the Kurdish organizations
operating in Europe in the 1990s, was viewed by many as the “legitimate
representative” of the Kurdish cause – the organization that was willing
Mechanisms of diaspora mobilization
81
to mobilize and use violence to press for Kurdish rights and autonomy.
The context here is important and a key to why this strategy worked.
A “radical” message or strategy will not be necessarily a winning
strategy in terms of mobilization success, but can be if an audience is
persuaded that it is the most legitimate strategy in a particular context.
In this case, a context of restricted membership in European societies,
severe repression in Turkey, and the perception that European foreign
policy enabled Turkish repression all worked to make the radical strategy of the PKK appear to offer a more effective pathway of resistance
than the more incrementalist approaches of the non-violent Kurdish
parties. At the same time, the PKK was skilled at altering its frames
over time to respond to changes in the context, as well as co-opting
the successful frames of other groups. For example, one sees a shift
over time from a more radical Marxist-Leninist message in the 1980s
and early 1990s to the emergence of a more human rights and political
and cultural autonomy frame in the mid to late 1990s – the PKK
effectively co-opted the successful elements of other organizations’
agendas into its own.
Resource mobilization: the diaspora as a source
of material support
Kurdish nationalist political entrepreneurs were able to draw upon the
politicized transnational networks that they activated in Europe as a
source of material support that the PKK used to build up its organizational structures within Europe and, in the 1990s, to fund the armed
conflict in southeastern Turkey. Much of the financial support for the
PKK was raised in Europe. The PKK harnessed material resources
from the Kurdish diaspora by collecting voluntary donations and
“taxes” of up to 20 percent of individual salaries and business profits
(Marcus 1993). The relationship between political mobilization,
resource mobilization, and impact on violent conflict is not necessarily
sequential – contributing to a cause can have a binding and mobilizing
effect on individuals, and the creation of a strong identity movement
can have an independent legitimizing effect on a conflict. Nevertheless,
the mobilization of material resources from the diaspora population is
one of the most direct impacts of the diaspora on the conflict.
Some sense of the scale of resource mobilization undertaken by the
PKK can be gleaned from official sources and police and intelligence
82
Fiona B. Adamson
reports. The German government estimated that the PKK collected
between Deutsche Mark (DM) 30 and 50 million annually via donations and racketeering from the Kurdish community in Germany in the
1990s – over DM 1.5 million annually in the city of Berlin alone.5
Rates of “donations” or “taxes” were set according to one’s financial
means and status, with the unemployed and asylum seekers expected
to pay DM 30–50 per month; gainfully employed members of
the community DM 100–300 per month, and rates of up to DM
3,000 a month for successful business owners.6 In the early days of
the conflict, money collected in Europe was reportedly transferred first
to a PKK office in Sweden, where it was consolidated from networks
across Europe and transferred directly to Abdullah Öcalan at his
headquarters in Damascus.
In the wake of Öcalan’s arrest in 1999, individual couriers reportedly transported revenues directly from Germany to southeastern
Turkey.7 The official ban of the PKK in Germany in 1993 apparently
led to an increase rather than decrease in the revenues that it collected
in Germany (Gunaratna 2001, 13–14). In addition to collecting
money from immigrant communities in Germany, the PKK also
actively collected revenues in a number of other European countries.
One Turkish source cited a British intelligence report that the PKK
collected £2.5 million in 1993 from immigrants and businesses in
England (Turkish Democracy Foundation 1997, 47).
Individual supporters of the PKK regularly provided voluntary
donations to the organization in amounts ranging from several thousand to over DM 10,000. Many of the large donations came from
well-established first-generation immigrants who managed to save a
substantial amount over the course of their stay in Germany, and
wished to exercise influence in the local Kurdish community by
making generous donations to Kurdish organizations (Imset 1992).
Many economic migrants during the first wave of migration from
Turkey in the 1960s and 1970s came from rural and economically
depressed regions of southeastern Turkey. Despite what were comparatively poor living conditions in Germany, most first-generation
5
6
7
Sources include Bundesamt für Verfassungsschutz 1996, 6; Landesamt für
Verfassungsschutz Berlin 1994; Der Spiegel, February 22, 1999, 35.
“Hilflos vor dem Terror,” Der Spiegel, March 25, 1996, 35 38, cited in Eccarius
Kelly 1999, 22.
Imset 1992, 156; Der Spiegel, February 22, 1999, 25.
Mechanisms of diaspora mobilization
83
migrants were members of German unions and earned middle-class
wages, which enabled them to save enough to start their own businesses or purchase land in Turkey. This meant that many in the first
generation were actually able to save substantial amounts of money
that could later be drawn upon for the funding of cultural and political
projects. Organized political fundraising activities sponsored by the
PKK were known to raise very large sums of money – one source
claims that the PKK was able to raise DM 100 each from 120,000
participants in a fundraising event in the Netherlands in 1998 (White
2001, 193).
In addition to voluntary contributions that were harnessed from
immigrant communities in Europe, an important source of revenue
for the PKK was extortion and protection money. This included protection money collected from Kurdish and Turkish businesses in Germany
and other European countries, and forced donations from individuals.
According to one source, 69 out of every 100 incidents of extortion that
occurred in Germany in 1994 were somehow connected with the PKK.8
The German government established a number of special commissions
in various cities throughout Germany to examine the problem of extortion and racketeering, in particular by the PKK.9 In the Netherlands, a
prime source of revenue was apparently extortion money obtained
from Kurdish-owned pizza parlors.10
Another source of revenue was the corruption of the asylum process.
A Swiss source argued that the PKK was collecting 20–60 percent of
the state support provided to Kurdish refugees who were waiting for
their asylum hearing, which resulted in annual revenues of millions of
Swiss Francs transferred from Basel, to Paris, and eventually to PKK
headquarters in Damascus (Helfer 1988, 48). Kurdish organizations in
a number of European countries, including Germany and the UK, were
responsible for certifying political asylum seekers as genuine. Both the
Turkish and German governments argued that the PKK, in addition to
receiving donations or other forms of payment from refugees and
asylum seekers, was also involved in controlling the market for forged
visas and passports, as well as human smuggling operations (Turkish
8
9
10
Statement by Rolph Tophoven, Director of the Terrorism Research Forum in
Germany, cited in Turkish Democracy Foundation 1997, 48.
Bundesamt für Verfassungsschutz 1996, 6.
Helle Bering, “Europe’s Kurdish Headache,” Washington Times, February
24, 1999.
84
Fiona B. Adamson
Democracy Foundation 1997, 48). The system was also reportedly used
as a means of sending PKK organizers from Turkey to Western Europe
and in engaging in human smuggling as a source of additional revenue
(Helfer 1988, 47; Dalman and Tabak 1995). The PKK, according to
some accounts, benefited directly from the asylum regime in three
respects: first, it charged refugees or others up to $3,000 to be smuggled
into Europe; second, it used this inflow of new refugees as a means of
bolstering the legitimacy of its claims regarding the extent of human
rights abuses in Turkey; and third, it mobilized and recruited these
refugees into European branches of the PKK, thus expanding its organizational structure and strength within the continent (Eccarius-Kelly
1999, 22).
Money raised by the PKK in Europe from donations and various
underground activities was, according to reports, used to purchase
arms originating from Germany, Iran, and Iraq (Imset 1991, 160). In
addition to purchasing arms and funding the PKK’s extensive organizational structure, the Kurdish diaspora in Western Europe was also
a source of PKK recruits for the conflict in the southeast. Figures on
the number of recruits that originated from Europe are difficult to
verify, but estimates put the number at 4,000 to 5,000. These recruits
often became militants after a period of training in Lebanon, and
others worked as “organizers, diplomats, technicians of various sorts”
(Bruinessen 1998, 45). Thus, the PKK, despite having a political support
base consisting of populations that were socially marginalized in both
Turkey and Western Europe, was successful in exploiting the resources
available in the diaspora as a means of funding the armed conflict in
southeastern Turkey.
Lobbying and persuasion: getting the Kurdish issue onto the
European political agenda
Perhaps as important as mobilizing political and material support in the
Kurdish diaspora in Europe, was the strategy of mobilizing political
support for the Kurdish cause in European capitals. This, in a sense, was
a version of the Keck and Sikkink “boomerang pattern” in which
blocked opportunities in one state (Turkey) created incentives to harness
the political clout of powerful actors in other states in a form of
“leverage politics.” Through a combination of agenda-setting, framing,
lobbying, and coalition-building, Kurdish political entrepreneurs were
Mechanisms of diaspora mobilization
85
largely successful in raising public awareness of the Kurdish issue in
Europe, making the human rights practices of Turkey more generally,
and the Kurdish issue in particular, one of the central focuses of discussions surrounding Turkey’s accession to the EU. Turkish state policy
towards the Kurds, and their military response to the PKK, came under
greater public scrutiny. The Kurdish cause gained a degree of legitimacy
in the broader public, and the Turkish state came under public and
diplomatic pressure.
A European-wide network of Kurdish information centers, for
example, provided points of contact for the press and politicians in
Germany and other states, disseminating alternative sources of information on the situation in Kurdistan. PKK information centers in
Germany issued press releases, maintained partnerships with German
press outlets, and organized debates with German politicians. Activities
designed to raise public awareness of the Kurdish conflict ranged from
forms of contentious politics such as mass demonstrations and festivals,
the taking over of buildings or other confrontational strategies such
as self-immolation, and hunger strikes. On a mundane level, posters,
placards, and graffiti were used as a means of ensuring that the Kurdish
conflict in Turkey became partially embedded within the fabric of
everyday life in Germany and other European states.
In addition to these “confrontational” strategies, Kurdish political
entrepreneurs also worked through institutional channels within
Europe (stergaard-Nielsen 2001). This included meeting with
German politicians, civil society groups, organizing educational events,
and pressing for Kurdish language classes in schools. The recognition
of identity claims by German authorities through such symbolic acts as
the sponsorship of cultural festivals or Kurdish language classes served
to legitimize the existence of a distinctive Kurdish identity – thus
strengthening the nationalist cause. This in turn provided the basis
for Kurdish political entrepreneurs to draw on human rights discourses
in order to make claims concerning the relationship between a Kurdish
political identity and Turkish state practices – and, by extension, the
nature of German foreign policy toward Turkey.
In addition to working through institutional channels, Kurdish
groups networked with leftist organizations in Germany and other
European states to organize events, including fact-finding missions to
southeastern Turkey, which were sometimes joined by German politicians. German pro-Kurdish solidarity groups traveled to Turkey and
86
Fiona B. Adamson
staged protests in the southeast on behalf of Kurdish villagers.11 Kurdish
political entrepreneurs in Germany also cultivated a sustained relationship with political parties – for example, a number of Kurdish politicians
at the local level in Berlin were members of the Green Party (Sayan and
Lötzer 1998, 25).
One of the strategies used by Kurdish nationalist political entrepreneurs was the establishment of an embryonic government-in-exile.
A Kurdistan parliament-in-exile was established in late 1994, and has
met intermittently in cities throughout Europe since that time. In 1999,
it regrouped as the Kurdish National Congress (KNC). Largely dominated by the PKK, and its political wing, the ERNK, its members were
elected from transnational constituencies and its activities include
holding press conferences, publishing literature on the Kurdish issue,
and dispatching delegations to third-party states, and to international
organizations, such as the United Nations, European Parliament,
and the Council of Europe. In 1995, the European Parliament recognized the PKK as a Kurdish liberation organization, arguing that it was
“what the FLN was for the Algerians and the ANC for South Africans”
(Watts 2004, 24). During the mid 1990s, its meetings attracted representatives from various human rights groups, NGOs, and political
parties. Members of the KNC established ties with national parliaments
in England, Poland, Italy, and Spain (Watts 2004, 25).
Kurdish political entrepreneurs drew on human rights norms and
discourses as a way of framing demands and articulating grievances
within Europe. By drawing on such discourses, they were able to
patch together transnational coalitions of support that brought together
civil society groups, non-governmental organizations, and actors
within various European governments. The networks of human rights
organizations and other organizations that make up a “transnational
advocacy network” on behalf of Kurdish human rights took on a
greater significance from 1999 onwards.
During this period, several major events occurred, including the
arrest of the PKK leader Abdullah Öcalan in Kenya following his
expulsion from Syria due to Turkish military pressure, his attempt
to obtain political asylum within Europe, and the opening of EU accession talks with Turkey, that raised international awareness of the
11
Reuters News Agency, “Turkey Holds Eleven Germans after Pro Kurdish
Protest,” April 17, 1995.
Mechanisms of diaspora mobilization
87
Kurdish issue and significantly altered the incentives and constraints
faced by Kurdish political entrepreneurs in Europe. These events
dramatically shifted the dynamics of the Kurdish conflict and ushered
in a new era defined by a weakened PKK who called for a ceasefire in the
conflict, but also by a wave of reforms and liberalization that accompanied the process of EU membership negotiations; the latter, in turn,
led to the removal of many of the restrictions on Kurdish language and
identity in Turkey.
The domestic landscape in Turkey changed further in 2001, when
the Islamic Justice and Development (AK) party came to power.
Despite reforms, the PKK abandoned its ceasefire in 2004. The conflict has continued to persist in recent years, but in a radically changed
context in which significant legal reforms have provided many of the
cultural, linguistic, and political rights to Kurds that had originally
been demanded by the PKK.
Conclusions
A growing body of research has shown that diaspora populations often
play a role in promoting or exacerbating violent conflict. However, it
has lacked detailed examination of the causal mechanisms at work in
the transnationalization of civil wars through processes of diaspora
mobilization. Rather than assuming diaspora support for conflict as
natural, or treating diasporas as unitary actors that have independent
effects on conflicts, this chapter has instead identified a number of
causal mechanisms that can be used to more closely and systematically
examine cases of diaspora mobilization in particular conflicts.
I have focused on a subset of five causal mechanisms and suggested
how they operated in the Kurdish-Turkish case. It is certainly possible
to identify many other relevant mechanisms and to engage in a greater
level of micro-analysis of each of them, as Bennett argues in his
contribution to this volume. However, my aim has been more modest:
To suggest the utility of a mechanism-based approach for understanding the relationship between diaspora mobilization and violent
conflict.
At the same time, it is important to acknowledge the limits of
focusing on the role of diaspora politics in violent conflicts. While such
an emphasis can illuminate some of the underlying dynamics, it does
not necessarily shed light on outcomes. Mobilization, fundraising, and
88
Fiona B. Adamson
lobbying may occur in a conflict, but such activities may not necessarily
be the only factors contributing to a particular conflict outcome.
The case of Kurdish nationalism and the PKK, for example, can be
contrasted with that of Kosovar nationalism and the Kosovo Liberation
Army. Both involved very similar causal mechanisms and processes of
diaspora political mobilization in Europe – including resource mobilization and lobbying – but resulted in very different outcomes in the same
year: a NATO military intervention in Kosovo on the one hand, and the
arrest of the leader of the PKK on the other. The explanation for these
different outcomes rests not in the processes of diaspora mobilization,
but in geopolitics: Turkey is a member of NATO and thus any military
intervention by the US or Europe would have been unthinkable.
|
4
Refugee militancy in exile and upon
return in Afghanistan and Rwanda*
kristian berg harpviken and sarah
kenyon lischer
Introduction
Over the past 25 years, it has been increasingly acknowledged that
refugees have some degree of political agency in exile and that this
agency has strong transnational dimensions. A main form of crossborder mobilization includes organized violence within displaced
populations with the aim of gaining political influence in the country
of origin. Popularly referred to as “refugee warriors,” the phenomenon
has been thoroughly analyzed (Lischer 2003; 2005; Stedman and
Tanner 2003; Terry 2002; Zolberg, Suhrke, and Aguayo 1986;
1989). In the original definition, refugee warrior communities possess
“a political leadership structure and armed sections engaged in warfare
for a political objective, be it to recapture the homeland, change the
regime, or secure a separate state” (Zolberg, Suhrke, and Aguayo
1989). Less attention has been devoted to the question of what
happens to militant refugees when they return (Harpviken 2009;
2010). What are the mechanisms by which returning refugees continue
to engage in organized violence? To what extent can post-return
militancy be explained by processes that started in exile (or even prior
to exile)?
Refugee militarization is generally part of a larger transnational
complex in which hosts and exiles have a history of interaction. Militarized refugee groups have the advantage of being able to shift
between the territories of a contender and host governments that are
either sympathetic to the rebels or unable to control their activities
* Research for this chapter was funded by the Research Council of Norway
(grant no. 185958). We are grateful for comments on earlier versions from Fiona
Adamson, Kristin Bakke, Jeff Checkel and Mark Naftalin, as well as constructive
feedback from members of the ‘Working Group on Transnational and
International Dimensions of Civil War,’ Centre for the Study of Civil War, Peace
Research Institute Oslo.
89
90
Kristian Berg Harpviken & Sarah Kenyon Lischer
(Salehyan and Gleditsch 2006). Transnational ties serve to channel
flows of personnel, arms, money, or knowledge, perhaps also fostering
political and diplomatic support. Particularly when rooted in common
networks that pre-date the period of exile, there is considerable potential for the growth of transnational organization.
In addition to exile-based interactions, transnational factors also play
a role in the reverse situation, when militarized exiles return home.
We argue that three primary mechanisms explain whether refugees
militarized in exile engage in violence upon return: (1) socialization;
(2) resource distribution; (3) security entrapment. With socialization,
we focus on transformative learning that fosters militant attitudes, often
occurring in exile, but, when effective, with long-term consequences. By
resource distribution, we refer specifically to the militant group’s ability
to selectively distribute various types of resources – money, positions,
land, business opportunities – to attract and sustain membership. Security entrapment describes a dynamic where a returning militant group
and its members perceive that the only way to withstand an existential
threat is to use force. These key mechanisms may interact sequentially,
as when recruits that were initially attracted by the group’s resources
over time are effectively socialized to share the group’s militant worldviews. The mechanisms may also operate concurrently, such as when
resource distribution and security entrapment reinforce the leaders’
commitment to militarization.
The United Nations High Commissioner for Refugees (UNHCR)
estimates that nearly 25 million refugees returned to their countries of
origin between 1990 and 2010 (UNHCR 2010a, 10). A proportion
of those returnees have been involved in political violence while in
exile, and some may do so upon return. Despite that, the potential for
militant return has received little attention, in part, as Richard Black
points out, because it lies “outside the core mandate of the key policy
organization involved with refugees – UNHCR – and after the period
in which the distinctiveness of the ‘refugee experience’ might be
expected to be most visible” (2006, 24). The neglect is similar across
a number of relevant analytical literatures. A case in point is international security studies, despite a concern with the military role
of refugees (Loescher 1992; Weiner 1992; 1996). The peacebuilding
literature most commonly sees return as an indicator of a successful
peace process, although at least one perceptive observer has questioned
this assumed relationship (Adelman 2002). Studies of disarmament,
Refugee militancy in exile and upon return
91
demobilization, and reintegration (DDR), a subfield at the interface of
security and peacebuilding, are similarly mute on what happens when
militarized refugees return to their homeland.
In the following section, we explain our mechanism-based approach
and elaborate on the three main mechanisms at play. Following
Checkel, we adopt Gerring’s definition of a mechanism as “the pathway
or process by which an effect is produced or an outcome is accomplished” (Gerring 2007b, 178). Hence, we establish a framework
that can elucidate the processes and interactions involved in returnee
violence. We then apply this framework to the cases of refugee mobilization and return in Afghanistan and Rwanda, both of which have had
large scale militarization of refugees. For each case, we then refine the
analysis of the mechanisms through an iterative process, going back and
forth between the framework and the empirics. In both Afghanistan
and Rwanda, we find that the socialization and resource distribution
mechanisms are most useful for understanding returnee-related violence, but with security entrapment playing a key role in instances where
there is a delayed remobilization following an initially peaceful return.
We take a historical perspective to uncover continuities (and ruptures)
in the way the mechanisms at hand play out in exile and upon return,
seeking to define their sequential interaction over time. We conclude
with further implications for the study of transnational mechanisms
in the context of refugee and returnee militarization.
Mechanisms of returnee violence
Socialization
Returnees who have undergone systematic socialization that breeds
radical or militant attitudes are more likely to engage in organized
violence for political purposes, particularly if the socialization
has fostered a vision of returning home as a means for rectifying
past injustices and building a new polity. Socialization involves
transformative learning, in the form of acquiring new skills, and also
new ways of thinking about the world. Socialization is often understood
as the outcome of interaction between individuals (what Bakke in her
chapter refers to as relational diffusion; see also Checkel, this volume);
yet institutions – schools and educational centers, organizations with
a major component of on-the-job training (i.e. armies), or cohesive
92
Kristian Berg Harpviken & Sarah Kenyon Lischer
solidarity groups – are a prerequisite to bring about societal transformation. We assume, in line with Checkel (2001), that people in novel
or uncertain environments, such as refugee camps, are most likely to
be susceptible to socialization.
Our main focus is on political and ideological socialization,
although we recognize that convictions may be solidified by developing
supporting skills. In military socialization, the acquisition of new skills
and the transformation of consciousness intertwine, as learning to fight
and kill is inextricably linked to changing perceptions of self, as well as
to justifications for exerting violence. Elisabeth Jean Wood offers a
valuable analysis of the transformation of social networks during war,
which includes her conception of military socialization (Wood 2008).
She explains that military recruits “have to be socialized in the use of
violence for group, not private, purposes.” She continues by noting
that “[t]raining and socialization to the armed group take place both
formally, through the immersion experience of ‘boot camp’. . . and
informally, through initiation rituals and hazing” (Wood 2008, 546).
Dara Kay Cohen, in her work on sexual violence during war, draws on
Bard O’Neill (O’Neill 2001) explaining that “socialization may be
achieved through the mutual hatred of an enemy group, as a result of
guiding ideology in which all the group members believe, or . . . a set of
group benefits that may be experienced from a group activity” (Cohen
2010, 24). While socialization can be conceived of in both rationalist
and constructivist terms (Checkel 2005), we contend that effective
socialization may breed militant commitment even among recruits
who initially joined for solely instrumental reasons (Wood 2010).
Socialization is reinforced by engaging in violence, as when fighters
from the refugee setting circulate back and forth across the border. The
likelihood of post-return violence will depend on whether militant
socialization continues, takes on new forms, or comes to a halt, as a
consequence of return. At its most effective, however, socialization
may foster attitudes that will persist over time, even in the absence
of organizational regularity or other forms of maintenance. For our
purposes, where the main interest is in post-return militancy, we expect
exile socialization to be essential.
The combination of formal and informal socialization is likely to
be particularly robust, and – particularly relevant for our purposes –
can bring about attitudinal change that survives the dissolution of
the fighting entity, rendering a latent militant potential that can be
Refugee militancy in exile and upon return
93
reactivated in the face of major upheavals. More broadly, the likelihood of violence increases when a large number of refugees sharing a
radical vision return together, as often occurs following major political
and military changes at home. Thus, in many conflict settlements,
rapid and voluntary return may carry with it the prospects for
destabilization.
The observable implications of socialization include changes in institutions and in discourse (both public and private). Seeing them as the
main harbingers of new worldviews, we want to capture both the
transformation of existing, and the formation of new, institutions. Data
on institutional innovation can be solicited from the broader case
literature (on the conflict and its specific groups), with the caveat
that the socialization efforts that take the form of direct interaction or
that are of a more informal character are often harder to document.
Furthermore, institutional capacity does not necessarily bring about
attitudinal change, and we therefore need to capture transformations
of the discourses on goals and motivations, as well as tactical choices
and justifications. The changes in the discourse can be documented
through interviews with group members or knowledgeable observers.
For institutional change, we will still examine what happens in exile,
both because institutions are often exile-specific (even though schools
and education centers are sometimes repatriated wholesale), and
because effective socialization outlives its institutions (and a return to
country of origin).
Resource distribution
In addition to socialization, an important mechanism related to
returnee violence is the ability to distribute valuable resources, both
during and after exile. In situations of civil war, political and military
organizations cannot succeed without ensuring individual contributions to the cause. The driving engine of the resource distribution
mechanism is the ability to offer penalties and rewards that target the
individual, often referred to as selective incentives (Olson 1965).
Resource mobilization theorists (McAdam, Tarrow, and Tilly 1996;
Tilly 1978) view robust access to resources as key to the success of
opposition movements of any sort (see also Bakke, this volume). In the
first instance, and in its concern to create incentives for individual
behavior, the mechanism operates with an assumption of – possibly
94
Kristian Berg Harpviken & Sarah Kenyon Lischer
bounded – rationality (Checkel 2005). Yet, at a deeper level, individual
commitment hinges on the ability of the organization to either build
upon, or gradually foster, a coherent network and a deeper sense of
shared purpose (Gates 2002). The sequence is one where the resource
distribution and socialization mechanisms combine, as members are
either coerced or paid to join, but where systematic socialization later
fosters loyalty and shared purpose.
In many refugee crises, large-scale returns take place as a response to
peace agreements or regime changes. Refugees may return when they
are attracted by either the prospect of controlling opportunities for, or
receiving the benefits of, patronage. The type and value of a resource
varies with context and can include: positions in government and the
military; claim to land and property; business advantages (such as
mineral concessions or a share in state-run enterprises); and control
over aid. The potential for offering selective incentives to followers
increases if the exile group has a stake in the new power structure.
Since the most attractive opportunities will fill quickly, a sense of
urgency may develop among returnees. Thus, there is an inherent
risk that those returning will use violence, or at least threaten to use
it, to maximize control over resources. It is important to note that
this mechanism applies primarily to those returnees who are already
enrolled in political-military entities, and have previous experience of
political violence.
What are the implications that one can observe when the resource
distribution mechanism is operating? We can observe the group’s
discourse on resources, what it does to gain control over them, and
whether resource distribution is converted into armed capacity. The
discourse over resources can be explicit, focusing on unfair distribution
or on the prospective rewards when gaining power; or it can be
more implicit, as when the setting up of a shadow government signals
the promise of future control over resources. Expert opinions and
interviews with members are the prime sources. Establishing the
resources made available to the group is not necessarily any easier, as
both suppliers and receivers have an interest in restricting information.
Again, the reporting by experts and interviews with key informants
are the best sources. Ultimately, however, our core interest is in whether
the distribution of resources affects the commitment of individuals
and the group’s capacity for violent action. Observable implications
here are whether increased commitment – in particular to militant or
Refugee militancy in exile and upon return
95
violent action – is linked to favorable access to various types of
resources or whether failure to comply with a group’s expectations
results in resources being withheld. Assessing whether the access to
resources, and the ability to distribute them selectively, affects individual
commitment and group capacity requires deep familiarity with the
trajectory of both the group and the conflict context.
Security entrapment
When the political settlement that initiates a return process does
not include a trustworthy and enforceable role for the group and its
members, it is highly likely that return will be followed by continued violent
engagement. Potential resolutions include internationally brokered powersharing deals (with or without international peacekeepers), domestically
negotiated agreements, and decisive military victories by one party. The
literature on civil war finds that a weakly enforced external compromise
increases the likelihood of violence (Walter 2003). In that situation,
returnees will fear for their safety, especially if they return as minorities
to a majority-held area. A military victory, on the other hand, can either
guarantee or erode protection for the returnees, depending on which
group emerges victorious.
To the extent that a post-return security dilemma reproduces security
threats that were associated with earlier flight, this further heightens
the likelihood of post-return violent engagement. In some instances,
perceptions of insecurity may develop over time during the reintegration
phase, leading to violence. For example, if a peace agreement starts to
unravel, previously militarized returnee groups may react to the security
vacuum by taking up arms. This may be important, not only for those
who are core members of the movement taking part in an “armed
return,” but also as a mechanism for remobilizing former loyalists
who have left the movement (who may or may not be earlier returnees
from a militarized exile). Security entrapment may also bring the principled bystanders (those who held back on joining) on board, as the
alternative is insecurity or even death.
What are the observable implications of security entrapment? The
starting point is the very nature of the security architecture, which in
most, but not all, cases of refugee return comes in the form of a peace
settlement. Is there a formula for power-sharing that accommodates
the fundamental security concerns of all major parties, and is there a
96
Kristian Berg Harpviken & Sarah Kenyon Lischer
credible external guarantor? The analyses by party representatives and
seasoned observers, both at the moment of transformation and in
retrospect, are interesting here. Second, and even more important,
there is the tracing of the main parties through the process of political
transition to detect the emergence of constellations in which the parties
with their members and larger constituencies will feel entrapped.
The sources are similar to the ones tapped to document the security
architecture, perhaps with more emphasis on insiders’ perspectives;
however, the empirical challenges may be greater now, both because
the processes may unfold within very short time periods and because
secondary sources may be in short supply. Finally, there are important
exacerbating factors that feed into the entrapment mechanism. Most
important, historic experiences of entrapment (prior to flight; in exile)
and their embodiment in narratives that may be activated by leaders
intent on mobilization can greatly stimulate the sense of entrapment,
and therefore the will to fight. Likewise, the reactions to dramatic
events, such as the death of a key politician (even if by natural causes),
are worth scrutinizing, as the ensuing analysis and reactions (or
absence thereof) are firm indicators of whether there is an evolving
sense of entrapment.
The security entrapment mechanism often interacts with other mechanisms. It connects with the socialization mechanism in that people
who have been trained to hate and trained to fight are more likely
to react violently if entrapped. In relation to resource distribution, it
can be seen as a functional equivalent to it (i.e. even if the organization
cannot give you money, it can give you security, which is the key
thing when threatened). And obviously, a group’s access to resources
has diminished (and will diminish further) in situations of security
entrapment.
Refugee militancy and violent returns in Afghanistan
While Afghanistan has become the model case of refugee militarization, there has been virtually no analysis of what happened to the
militarized refugees upon return (but see Harpviken 2009; 2010). The
militarization of large parts of the refugee population in Pakistan, and
its importance for the evolution of the conflict, has been thoroughly
demonstrated, not the least in the pioneering work of Zolberg,
Suhrke, and Aguayo (1986; 1989). In this case study, we will focus
Refugee militancy in exile and upon return
97
on the mechanisms by which some returnees came to engage in
violence upon return, while others did not.1
Afghanistan has been characterized by armed conflict since the late
1970s, accompanied by the largest refugee movements in recent times,
with an estimated peak at over 6 million refugees in Iran and Pakistan
by 1990 (UNHCR 2001). Identifying the mechanisms at work, our
point of departure will be the militarization of refugees in Pakistan,
from 1978 to 1992, followed by an examination of refugee return
in the transition from the People’s Democratic Party of Afghanistan
to the so-called Mujahedin government (a resistance coalition) in 1992,
and the Taliban challenge to the internationally supported Karzai
government in the aftermath of the 2001 US-led intervention. We
will see that socialization is important to exile-based militancy, but
with varying impact upon return, depending on whether the resource
distribution and security entrapment mechanisms are at work, and that
understanding the interaction between mechanisms is critical.
The emergence of a state-in-exile (1978–1992)
What would become the core of the Pakistan-based resistance of
the 1980s originated in the ideological socialization of Islamist study
circles at Kabul University in the 1960s and early 1970s. Politically
radical, with ties to the international Muslim Brotherhood, this was a
small group with little resonance in the larger Afghan population. After
the kingdom was converted into a republic through the 1973 coup, a
group of some 1,000 Afghan Islamists were welcomed by Pakistan,
which saw them as a potential asset against a troublesome neighbor
(Roy 1986, 74–79). At the turn of the decade, this small group of
Pakistan-based radicals would become the pool from which the host
country’s intelligence service selected the leaders of the refugee-based
insurgency and its state-in-exile (Rubin 1991). In April 1978, the
Soviet-oriented People’s Democratic Party of Afghanistan (PDPA) took
1
This case study is based on an examination of the relevant secondary literature
(including academic, journalistic, and biographic accounts), as well as 20
interviews with key informants (conducted by Harpviken in 2008 and 2009), and
110 interviews with Afghans who have been engaged with a variety of political
military organizations (conducted by a group of Afghan surveyors in 2009). The
study also builds on previous work by Harpviken on political mobilization in
Afghanistan (Harpviken 2009).
98
Kristian Berg Harpviken & Sarah Kenyon Lischer
power in a coup, with a forceful reform agenda that immediately
sparked local uprisings. When the Soviet Union intervened in December
1979, Afghanistan became a hotspot of the Cold War, and within
two years, the refugee population was an estimated 3 million.
Resource distribution
For the Afghan refugees in Pakistan, we see the resource distribution
mechanism at work in its crudest form. The early resistance had largely
been led by local notables, often instrumental in facilitating flight for
their followers, intent on continuing warfare from the relative safety of
exile (Harpviken 2009, 55; Suhrke and Klink 1987). In exile, the
embryonic political structure among Afghan refugees transformed
quickly. The host government effectively decided on a set of seven
groups that would represent the Afghan resistance. Each group was
led by one of the long-standing exiles. Intent on supporting the Afghan
resistance, but concerned about being drawn into the war, Pakistan
sought to maximize influence over all sorts of international assistance,
ranging from military supplies to humanitarian aid (Fielden 1998;
Schöch 2008).
A particularly effective measure was to link refugee aid and politics.
Refugee status was made conditional on membership in one of the
political parties, with the net result that the absolute bulk of refugees
became party members (Amstutz 1986, 229). The majority of the
refugees were settled in camps, often under the control of a specific
party. The parties had considerable influence over the distribution of
aid in the camps (Centlivres and Centlivres-Demont 1988). Traditional
notables were often marginalized, as a younger generation with a
different skill set – and a different political orientation – took over.
The control that parties gained over the distribution of all sorts of
refugee assistance was key to the political enrolment of refugees, but
also key to keeping a tab on refugees over time.
Male refugees, in return, were expected to be at the service of the
party. A common pattern was that fighters would spend up to six
months at a military base in Afghanistan, then come back to Pakistan
for a similar period, stock up on supplies, perhaps get refresher
training, rest and spend time with family. Truly mobile military groups
were rare, so most would be fighting in their area of origin (Roy 1986,
176–183). Despite the ties to local residents, however, there were
clear limits to the involvement of military groups in the day-to-day life
Refugee militancy in exile and upon return
99
of the society (Dorronsoro 2005, 212). This led to a relative seclusion;
the checks and balances of civil society lost out to the influence of the
political parties, which played a central role in most areas of life.2
Socialization
The political parties also emphasized socialization. While the pioneer
migrants were largely socialized prior to migration, the bulk of the
post-1978 refugees were – often reluctantly – brought into the orbit of
the political parties through resource distribution, with socialization
following suit. Socialization in this context would imply education
(both conventional and religious schooling), military training, and
lastly, combat socialization. The key, in relation to the prospects for
post-return violence, is that sustained education efforts in exile –
reinforced during combat tours across the border – fostered a radical
and militant vision. Post-return violence, then, is the delayed effect of
socialization that has produced robust convictions. The refugee context
for Afghans in Pakistan was well suited for effective socialization. The
refugees faced great uncertainty. They lived in urban centers or in
densely populated camps (often in remote areas), with the latter being
most dependent on the parties.
Conventional schooling was considered important by the Afghan
parties, particularly by the Islamist ones. In his doctoral thesis, Farhad
Rastegar (1991) offers a unique understanding of Afghan refugees
in Pakistan. Rastegar contrasts Hezb-e Islami with Jamiat-e Islami,
the two main Islamist parties. While Jamiat advocated a bottom-up
political transformation as the ultimate outcome of individual change,
Hezb pursued a top-down strategy whereby institutional change will
force social and individual change. To bring about the change of the
polity, however, requires capacity that can only be built in the schools,
in the form of a critical mass committed to revolutionary violence
(Rastegar 1991, 119–127). Schools, says Rastegar (1991, 261), “provide
movements with both a socially desired service outlet, and an easily
reproducible mobilizational structure.” As we shall see, upon their
1992 return, both Jamiat and Hezb, with their weight on exile socialization, proved capable of engaging in sustained large-scale fighting.
2
In the aftermath of the 1987 Geneva Accords between Pakistan and the Soviet
Union, the resistance parties issued threats against what they saw as an act of
treason, seeing return as contrary to the holy war against infidels in which they
were engaged (Rizvi 1990).
100
Kristian Berg Harpviken & Sarah Kenyon Lischer
Other groups, particularly those of a more fundamentalist politicalIslamic orientation, emphasized religious schools. This was particularly
the case for the Harakat-e Inqelab, led by Mohammad Nabi. The party
was rooted in the informal networks of Sunni Islamic scholars, in which
particular personalities with their own madrasas and entourage were
the building blocks (Borchgrevink 2010). The rapidly expanding
madrasa sector in Pakistan, catering both to Afghan refugees and
Pakistani nationals, contributed significantly to a hardening of political
sentiments. But the role of the madrasas, as is often alleged, was not to
build military competence – that was taken care of elsewhere – but
rather to foster a new political consciousness (Zahab and Roy 2004).
In that sense, the madrasas played a role similar to that of ordinary
schools (Fair 2009). Both schools and madrasas, particularly when run
by a political group, not only fostered militant worldviews, but also
served as effective conduits for recruits to the military branch of the
organization, where they would also develop their skills as fighters.
While many of the parties had some capacity to train their own
personnel, the bulk of the military training was arranged by Pakistan’s
security apparatus. Mohammad Yousaf headed the Afghan Bureau of
Pakistan’s Inter-Services Intelligence (ISI) from 1983 to 1987, and has
described the training:
At the end of 1983 we were operating two camps in Pakistan, each with a
capacity of 200 trainees. By mid-1984 we were putting over 1000 a month
through the system, and by 1987 we had seven camps operating simultaneously. (Yousaf and Adkin 1992, 117)
The overwhelming majority of the trainees had a refugee connection in
that their families resided in Pakistan.
The close interaction between political consciousness raising and
military training – even if it was kept physically separate – is interesting.
In Olivier Roy’s words, “politicization has arisen through militarization
and not as a result of implanting political parties or ideological
discourse” (Roy 1989, 45). Military socialization – at exile training
centers and in combat – became the fundamental carrier for ideological
reorientation. It is not so much an overarching ideological project that
necessitates militancy as it is a military engagement that brings with it
attitudinal political change. With militancy having become as much the
end, as a means to an end, it becomes difficult to foster reorientation
and compromise. Militant attitudes alone – fostered through the various
Refugee militancy in exile and upon return
101
forms of schooling – are one thing; however, when that is followed by
practical military training – and an ability to practice the newly gained
skills – militancy becomes particularly robust.
Regime change and civil war (1992–2001)
When, in April 1992, the remains of the PDPA regime collapsed,
massive repatriation followed swiftly. The seven exile parties negotiated
a deal, and their leaders moved to Kabul in the company of their armed
entourages, each taking control in various parts of the city. Government
ministries were quickly converted into the armed fortifications of the
party who had the interim minister. The outgoing regime had split into
factions that merged with various resistance fronts. A particular case
was Rashid Dostum’s Junbesh-e Milli, a militia whose changing of sides
was pivotal to the government collapse. Within a few weeks, Kabul was
thrown into massive violence, first with Sayyaf’s Ittehad forces shelling
Shia majority areas in West Kabul (Harpviken 1996, 113). Not long
after, Hezb-e Islami, whose leader was given the prime ministership in
the treaty, started rocket attacks across Kabul (Rasanayagam 2003,
142–143). Over the next years, Kabul became a battleground in which
the former resistance parties fought each other with great intensity
and with massive civilian costs. The main parties to the armed politics
playing out in Kabul were the two Islamist groups built up in Pakistani
exile, plus Hezb-e Wahdat (an alliance of Shia parties with offices in
Iran) and Dostum’s militia-based group. That other exile groups did
not display an appetite for engaging in the fighting makes comparison
interesting.
Socialization
When peace broke down in 1992, to what extent did the socialization
of exile manifest itself in sustained violent engagement? One of the
main contenders, Dostum’s Junbesh-e Milli, was not bred in exile.
The other groups were not exclusively exile-based, and they had all
incorporated elements from the army of the previous regime. Nonetheless, the one group that proved to be the least reconciliatory, and that
has to take a major part of the blame for the spiraling violence in Kabul
from 1992 – Hekmatyar’s Hezb – was also the group that had the most
systematic approach to militant socialization. While the disconnect
between the Hezb and existent social structures helps explain its failure
102
Kristian Berg Harpviken & Sarah Kenyon Lischer
to mobilize broadly (Harpviken 1997), the same seclusion also permitted
effective socialization around radical political visions. There were local
groups that peeled off, and there was the inclusion of army elements, but
the Hezb leadership managed to sustain its capacity throughout a fight
that most Afghans saw as totally illegitimate (Roy 1995, 63–75).
Resource distribution
The resource distribution that works so well to instill and maintain
organizational loyalty in exile is hardly reproducible upon return,
when the sources of support are changing, the expectations of armed
groups and their constituents are altered, and the conditions by which
a group monitors members and fosters dependence are less favorable.
The 1992 transition did not abide by the logic of a peace treaty.
Rather, gaining control over territory (and elements of the fragmenting
state apparatus) became decisive for what political claims any group
could expect to successfully make. Rapid mobilization – and therefore
rapid return – was critical. Fieldwork findings from 1992 to 1994
(Harpviken 1996) are reconfirmed by the interviews conducted for
this project in 2010. For those who have served the party in, or from,
exile there is an instinctive expectation of a fair share in the spoils of
peace, with real prospects of getting access to good positions with the
new administration. This is reflected in the peak of returns just after
the regime change: 1,274,000 returnees were registered by UNHCR
in 1992, less than 300,000 the following year (UNHCR 2003).
Over time, the willingness to take risks for the party diminishes if
resource distribution fails. Hezb-e Islami, arguably the party with the
strongest party loyalty, did lose some of its commanders as international
funding dried up after 1992 (Giustozzi 2009, 70–71). Yet, the ability to
maintain its military engagement despite a severing of resource distribution capabilities indicates that strong socialization can in large measure
render resource distribution unimportant (Humphreys and Weinstein
2006). In contrast, those groups who had a poor socialization record
and lacked access and ability to distribute resources quickly fragmented
(Giustozzi 2009, 71–73; Rubin 2002, 256–264).
Security entrapment
The early returners expected to be part of an orderly transition to a
new regime, not to a new round of civil war. The fighting, particularly
as it played out in the capital, created a situation where one’s security
Refugee militancy in exile and upon return
103
depended entirely on protection from one of the armed groups. The
forces of the various parties carved out their own territories, in large
measure associated with a particular ethnicity: Jamiat being dominantly Tajik, Hezb-e Islami mainly Pashtun, Jumbesh heavily Uzbek and
Hezb-e Wahdat almost exclusively Hazara. The escalating insecurity
spiral in Kabul from April 1992 onwards placed anyone at risk, and
those who had an affiliation with any political group – indeed the case
for almost all returnees – found themselves compelled to seek protection
there (Harpviken 1996, 116–119). Such protection, naturally, would
come at the price of not only loyalty, but also tangible contribution to
the armed struggle. An alternative was to escape; perhaps as many as
1.5 million people did. These were mostly those who resided in Kabul
under the old regime, and who therefore lacked the connections that
were necessary to solicit the protection of the new rulers.
Meanwhile, a great deal of those who had fought from exile in
Pakistan during the 1980s had disassociated themselves from the
national power struggle. For them, the holy war – the jihad – was
won (Roy 1995, 68–72). This was the common pattern both among
the so-called traditionalists and the fundamentalists. These were
groups that had largely lost out in the transition from the PDPA, even
though they were maintaining some posts in the government structure.
They also stood at the sidelines when the battles for Kabul played out.
Why? While these groups were centrally placed in the madrasa sector
in exile, their religious education was more traditional, less ideologized,
and, at least in part, disconnected from military struggle. The fundamentalist groups had never gained the same degree of control over
resource distribution in exile as had the Islamists. One could still have
expected that their claim to possible influence would stimulate military
action in the context of return, but while the leaders did return to Kabul
with their trusted armed men, they never threw themselves into the
new civil war. Ultimately, these were groups whose roots were in the
rural areas; as a result, the consistent, enduring, security entrapment
that ignited military action within other groups rarely occurred.
The coming to power of the Taliban, through a massive armed
campaign from 1994 to 1996, will not be analyzed in any detail here.
Yet, it is worth noting that the Taliban in large measure incorporated
the constituency of the parties that held a more traditionalist or fundamentalist orientation – the very parties who had not taken part in
the fight over Kabul post-1992. Undoubtedly, a religiously founded
104
Kristian Berg Harpviken & Sarah Kenyon Lischer
political vision brought about by socialization in exile was instrumental to the Taliban mobilization, yet its ability to sustain itself
was contingent on resource distribution, with Pakistan being a willing
supplier and enabler. Mobilizing as much within the refugee population as within Afghanistan itself, the Pakistani dual role of refugee host
and source of support was key to Taliban expansion.
International intervention and armed remobilization (2001–)
The armed resurgence of the Taliban, following its ousting by the
US-led intervention from 2001, took two or three years to gain
momentum. In an intervention in which the ground forces were exclusively Afghan (with a few international advisors and special agents),
the remobilization of fighters in exile played a prominent role (Rashid
2008). Examples of this are Abdul Haq, a legendary resistance commander who was captured and killed by the Taliban in the Jalalabad
area, and Hamid Karzai, the president-to-be who went across to
Kandahar with his own men and a US support team (Barfield 2010,
288–290; Schroen 2005, 273–278). Also, the so-called Northern
Alliance, and the strongest party within it – Jamiat-e Islami which by
2001 maintained control over perhaps 10 percent of the territory (and the
country’s UN seat) – drew heavily upon exiles. While it is clear that the
refugee return component of the events in 2001 and beyond was significant, the focus here shall be on the Taliban resurgence that followed what
at first appeared to be a decisive and widely popular intervention.
The literature on the Taliban resurgence (Giustozzi 2007; Rashid
2008) discusses the importance of Pakistani support – including access
to sanctuary – but does not explore its transnational dimensions (but
see Franco 2009; Stenersen 2009). Yet, there is little doubt that the
post-2001 Taliban mobilization was firmly related to the migration
and refugeehood that had characterized the past 20 years, and that
recruitment among recent returnees, as well as among exile Afghans in
Pakistan, was essential to the organization. Patterns of movement
across the border, however, had become increasingly complex. What
is often referred to as the neo-Taliban was neither a purely refugeebased insurgency nor a purely returnee-based one, but it was certainly
one that drew on refugees and returnees alike, and not least on decades
of intensive mobility that had fostered dense networks of interaction
across the border.
Refugee militancy in exile and upon return
105
Socialization
What was the role of socialization in the resurgence? Most importantly, the combination of exile political training with battlefield
socialization did breed a sense of common purpose. In the immediate
aftermath of September 11, 2001, the sense of common purpose was
severely weakened, as a large share of former Taliban sought distance
from an ideologized religio-political platform – rooted in Wahabist
Islamic thinking – that many felt was foreign. The Taliban leadership
responded to this by modifying their political message, in Thomas
Barfield’s words, portraying “themselves less as Muslim zealots and
more as God-fearing nationalists seeking to expel infidel foreigners
from the country” (Barfield 2010, 327). This resonated better with
the dominant mode of thinking in the refugee environments in Pakistan
and in most madrasas on both sides of the border.
Triggered by an international military campaign that focused on the
“war-on-terror” – at times with severe consequences for civilians,
while everyday security under the new government was declining
severely – the Taliban’s message resonated fairly well with ingrained
attitudes. The pool of recruits that would over time swell the ranks of
the new Taliban was easy to miss in 2002 and 2003, since they were
either getting on with their ordinary lives in Afghan society, or spent
most of their time in the refugee environments across the border;
however, they had both the right ideological outlook and the necessary
military practice. Once the sentiment within the mainly rural Pashtun
population in Afghanistan’s south and east started to turn against the
Karzai government and the international forces, the Taliban tapped
into an attitudinal base that gave it strong momentum.
Resource distribution
Resource distribution was not a key element of Taliban remobilization.
Historically, the control over resources had been important to draw
Afghan refugees to the various political parties, and the Taliban built
on this. The resource distribution that the Taliban represented was
limited to basic security and Islamic justice, and, at the personal level,
recognition and positions of power. The Taliban was never a welfare
organization; its vision of the state’s responsibilities was very narrow.
Neither, despite the common reference to non-ideological, easy to
convert, “ten-dollar-a-day” Taliban (e.g. Miliband 2009), was it
an army of guns for hire. Most important, though, was the relative
106
Kristian Berg Harpviken & Sarah Kenyon Lischer
promise that stemmed from the fact that great expectations of aid
and economic development in the wake of the US-led intervention
were not met. The impression early after 2001 was that the Pashtun
south received far less than its fair share of international assistance
(International Crisis Group 2003). With deteriorating security, neither
the government apparatus nor the internationals managed to change
that impression, even though there was significant economic growth in
more shady trades, drugs included. So, while the Taliban did not find
themselves in a position where they could use resource control to
attract followers, this was not as significant an impediment as it could
have been, given that the government and its supporters failed to offer
an alternative.
Security entrapment
The key to Taliban resurgence is security entrapment. The intervention
was, in the early period, widely popular among Afghans, including
a large share of those who had sided with the Taliban. Facing a
combination of supreme air power and Afghan ground forces, the
Taliban regime collapsed, and within a couple of months, the war
was effectively over. As the main leadership fled, often to Pakistan,
the rank and file Taliban’s fighters effectively demobilized, overwhelmingly with the intention not to re-engage in war (van Bijlert
2009; Strick, van Linschoten, and Kuehn 2011). Over the next couple
of years, however, an increasing number of former Taliban supporters
reconsidered their retirement ambitions, joined by a new younger
generation (Nixon 2011).
We are focusing here on the insurgency in the south of the country,
where it started first. The international forces – a continuation of
the intervention, Operation Enduring Freedom – were not set up to
offer security for citizens. Rather, they were focused on tracking down
al-Qaeda and high level Taliban, with a net effect that was negative.
Both non-aligned citizens and demobilized Taliban fighters had to bear
the cost of collateral damage to life and property. The UN-mandated
International Security Assistance Force (ISAF) at first had a remit
limited to the capital Kabul, but expanded to the south in mid 2006.
By then, the resurgence had already gained significant speed, and
rather than attempting to change the dynamics, the ISAF forces
adopted an offensive anti-Taliban strategy that came with significant
costs in terms of insecurity and damage.
Refugee militancy in exile and upon return
107
Also the government, whose structure was in large measure built
on people whom the Taliban had driven from power in the 1990s,
constituted a threat rather than a source of protection. While there were
virtually no ordinary armed forces, there were groups under the command of local officials and other strongmen, engaged in a variety of
pursuits from local power struggles to subcontracting for international
forces, but again with a net negative effect on local citizens’ security
(Gopal 2010; Moghaddam et al. 2008). Once insecurity triggered armed
protest, and neither the international forces nor the government proved
capable of reacting effectively, a spiraling insecurity dynamic unfolded in
which the armed responses by various actors were mutually reinforcing.
Distancing oneself from the conflict became more difficult by the day,
and for many, there was no alternative to joining the Taliban.
What are the overarching patterns in the Afghan case? For the
Afghan refugees in Pakistan in the 1980s, a rich resource supply from
states and aid agencies was placed at the disposal of political parties
handpicked by the host state. These parties used the resources effectively to draw refugees into their institutions of socialization. The
parties ran effective socialization programs in exile, combined with
battle socialization in extended periods of cross-border mobilization.
With the transition in 1992, the promise of government positions and
other privileges was immediately important, but as fighting broke
out, was replaced by a genuine sense of security entrapment, in which
nobody but militant groups could offer critical security to their potential followers. Similarly, for the Taliban remobilization post-2001,
entrapment seems to have been the dominant mechanism at play.
Refugee return and political violence in Rwanda
The Rwandan conflict represents the most complex and violent
instance of refugee return in Africa since 1990. Political violence
involving refugee populations has influenced Rwandan politics since
at least the mid twentieth century. During the period of Belgian
colonialism, the minority Tutsi enjoyed favored status relative to the
Hutu. This power balance reversed in the late 1950s when the Belgians
withdrew support for the Tutsi king. Following independence in 1961,
at least 200,000 Tutsi civilians fled to Uganda, Tanzania, and Zaire.
Thirty years later, a Tutsi-led exile army invaded Rwanda from
its base in Uganda with the goal of toppling the oppressive Hutu
108
Kristian Berg Harpviken & Sarah Kenyon Lischer
government. The Rwandan Patriotic Army (RPA) was led by refugees
(or their descendants) who had fled decades earlier. In the context
of civil war and political upheaval, extremist Hutu militants, soldiers,
and ordinary citizens killed up to 800,000 Tutsi and moderate Hutu
in the spring of 1994. As the RPA approached the capital city, the
government army and extremist militias fled to eastern Zaire, along
with over a million Hutu civilians. Tens of thousands of Hutu militants
infiltrated the refugee camps, intimidated refugees, stole food aid,
and launched cross-border attacks against Rwanda. In 1996, the
Rwandan Patriotic Front (RPF) government responded by attacking
Zaire and forcing hundreds of thousands of refugees back over the
border to Rwanda. The government, dominated by the Tutsi returnees
from Uganda, prevented the previously militarized Hutu returnees
from regaining political power.
The mechanisms of socialization and resource distribution, within the
context of security entrapment, help explain that violence. To trace
the mechanisms at work requires examining Rwandan politics during
the 30-year exile of Tutsi refugees in Uganda, the civil war and genocide
of 1990–1994, and the reintegration period from 1994–2010. The
relevant mechanisms began germinating decades before the genocide
and have continued to shape politics in the years following. By examining those time frames, the remainder of this case study isolates the
relevant mechanisms.3
Tutsi refugees in Uganda, 1959–1990
Socialization
As discussed earlier in the chapter, socialization involves learning new
skills and ideas which can be transmitted through formal institutions
or informal networks. The socialization mechanism is not inherently
militaristic. Such influences can either encourage or discourage violence,
3
The following study is based on field research conducted in Rwanda in April
May 2009. About two dozen interviews with Rwandan and international staff
of local NGOs, religious organizations, community reconciliation and
reintegration programs, United Nations and other international organizations,
US government agencies, and returned refugees were conducted. It also draws
upon secondary sources, interviews with scholars with extensive field research
experience, and resettled refugees in the US. Due to the severe restrictions on
freedom of speech in Rwanda, most informants agreed to speak only under strict
conditions of confidentiality.
Refugee militancy in exile and upon return
109
depending on the social environment. The power of socialization may
be particularly strong among refugee populations, considering the
inhabitants’ limited sources of information, geographical isolation and
anxiety. In her discussion of socialization in pre-genocide Rwanda, Lee
Ann Fujii cites a Rwandan whose own understanding of what affects
people’s behavior “assumes that people are influenced by those around
them and thus will inevitably take on the beliefs and attitudes of others.
According to this logic, the social environment is powerfully transformative” (2009, 101). This is especially likely in a context of insecurity and resource scarcity.
Observable implications of socialization in exile include the transformation of refugee-based institutions from peaceful to explicitly
militaristic. The successful RPF invasion had its roots in years of
organized exile activity, which included ties with Rwandans in distant
countries. As discussed earlier in the chapter, such conditions increase
receptivity to militant socialization. The experience of persecution
both in Rwanda and in Uganda encouraged many refugees to join Tutsi
refugee organizations, such as the Rwandan Refugees Welfare Association
and, later, the Rwandan Alliance for National Unity (RANU).
These groups created community and socialized members in a way that
facilitated a transition from peaceful community building to support
for militarized return. For example, during the mid 1980s, RANU sent
refugees from Kenya to Uganda for military training (Kuperman
2004, 66). Leaders also benefited from the ability to distribute resources
received from the diaspora. These resources provided material encouragement once the militarized socialization programs took root. Prunier
reports that the nascent RPF army received financial support, and
even recruits, from the Tutsi diaspora (Prunier 1998, 124, 131–132).
Thus, the decades of socialization facilitated the leaders’ later efforts
to form an exile army.
Security entrapment
Socialization was not the only contribution to militarization, however.
Security entrapment and resource distribution mechanisms interacted
with socialization in exile. As the Ugandan exile persisted, local resentment intensified due to a perception of preferential treatment by the
government and the United Nations. In response, the Ugandan government attempted to confine the refugees to camps, leading to an isolated
and impoverished existence (Mamdani 2001, 164–165). Mamdani
110
Kristian Berg Harpviken & Sarah Kenyon Lischer
explains that “a refugee self-consciousness developed first and
foremost in response to anti-refugee prejudice promoted by the state
and shared by many in the society at large” (Mamdani 2001, 165).
In the late 1960s Ugandan President Milton Obote cracked down on
Tutsi economic activity (van der Meeren 1996, 261). Idi Amin then
took power in the 1970s and recruited a few Tutsi refugees to fight
with him, although eventually the relationship soured. Obote regained
power in 1979 and increased violent scapegoating of the refugees to
the extent that Mamdani labels it a “pogrom” (Mamdani 2001, 168).
That oppression increased refugees’ fears for their survival and led
the Tutsi to support Yoweri Museveni’s rebel forces, which were
victorious in 1986. By that time, around 3,000 of Museveni’s 14,000
rebels were Rwandan Tutsi exiles (Kuperman 2004, 66). During their
time with the Ugandan rebels, the exiles that later formed the RPF were
socialized under tight military discipline and learned basic guerrilla
warfare skills. Mamdani observes that: “The electric example of a
home-grown guerrilla movement that had defeated an internationally
recognized government without substantial external support was not
lost on RPF leaders” (Mamdani 2001, 173). By the late 1980s the
refugees’ situation in Uganda became increasingly untenable, however,
as Museveni realized that the Tutsi constituted a political liability.
Following anti-refugee violence, “the guerrillas harvested youthful
recruits from the victimized population” (Mamdani 2001, 170). In
December 1987, the Rwandan refugee organization altered its focus
to emphasize the goal of return to Rwanda, by force if necessary, and
renamed itself the Rwandan Patriotic Front (Lemarchand 2001, 12).
Prunier points out an observable implication of the socialization spurred
by this security entrapment: “the result of these changes occurring at
the top was that many people lower down decided to turn to the RPF
and its radical project of reconquest” (Prunier 1998, 127).
Resource distribution
As discussed earlier, the ability to provide resources (or withhold them)
offers significant leverage to military groups in exile. Upon return,
those groups who maintain that capability will be more likely to
mobilize militarily. While in exile, Rwandan refugees had four main
resource streams – the Ugandan government, international organizations, diaspora funding, and their own income-generating activities
such as farming. Humanitarian assistance provided one important
Refugee militancy in exile and upon return
111
source of resources, even including UNHCR scholarships for youth.
Such aid, however, created resentment from the local Ugandan population and stoked violence against the refugees (Mamdani 2001: 165).
The Rwandan Tutsi exile organizations in Uganda gained recruits
whenever the resources provided by the Ugandan government and
international organizations decreased and the relative importance of
diaspora funding increased.
Political and economic oppression by the Ugandan government
solidified the militant resolve of the RPF and its search for additional
resources. As the Uganda-based exiles coalesced, the “new militants
from the exiled left-wing intellectuals in Europe and North America”
increased their support (Prunier 1998, 128). RPF adherents viewed the
diaspora-funded organization as a better resource relative to the other
dwindling options. Kuperman (2004, 67) reports that the RPF received
up to $1 million per year from the diaspora. In addition, the RPF could
promise its members future opportunities for land and employment
if they toppled the Rwandan government. As Mamdani (2001, 167)
observes, “the question of entitlement becomes acute at times of political crisis.” In the context of uncertainty about the future in Uganda, the
socialization and resource distribution mechanisms complemented each
other to ensure the success of the RPF invasion and its consolidation of
power. RPF socialization strengthened cohesion and the promise of
resources increased that solidarity upon return.
As the various push and pull factors fell into place, the RPF army
crossed the border into northern Rwanda in October 1990. After initial
setbacks, the invaders made significant progress, though at a high cost
to civilian life – both Hutu and Tutsi (often killed in retaliatory attacks by
local forces). The RPF secured territory fairly easily, but it failed
to establish political dominance. As the rebels advanced, the terrified
population fled wholesale, leaving the RPF in control of abandoned
towns, villages, and farms. Military socialization in exile was not sufficient to enable the RPF to build a government in the conquered territory.
Civil war and genocide, 1990–1994
Security entrapment
As the RPF gained a military advantage, the international community
pressured all the combatants to negotiate. This pressure reflected
Western norms favoring power-sharing and ethnic diversity, as well
112
Kristian Berg Harpviken & Sarah Kenyon Lischer
as a reluctance to commit military resources to the region. Western
states and the United Nations cajoled the parties to sign the Arusha
Peace Agreement in Arusha, Tanzania. The Agreement did not reflect
the preferences of either the Tutsi rebels or the government of President
Habyarimana, and left both in a situation of insecurity. The RPF
agreed to withdraw its forces to specified points and the government
promised to embark on democratization, including the integration
of the Tutsi forces into the army and government (Arusha Peace
Agreement 1993). The RPF had lost its stronghold in Uganda and,
thus, its fall-back option. The weak United Nations commitment
exacerbated the security entrapment since the agreement was not
enforceable. The fears of the Tutsi were magnified by their historical
experience of insecurity in Uganda. The leaders could call upon the
narrative of past oppression to mobilize followers.
The uncertainty about the future peaked after the president’s plane was
mysteriously shot down in April 1994. At this point, the Hutu extremists
unleashed the genocide and the RPF took up arms. Surprisingly, the
Rwandan government put up little resistance to the Tutsi rebels, seeming
to focus its energy on committing genocide rather than defeating the wellorganized and armed RPF. As the RPF advanced, fleeing Hutu militants
herded nearly 2 million Hutu civilians to neighboring states. Over one
million Hutu huddled in IDP camps. The rebels captured the capital,
Kigali, in July 1994 (Lischer 2005, 41–54; Prunier 1998).
Socialization
In tracing the story of the civil war and genocide, one sees that the
socialization process that unfolded in exile continued to affect returnee
violence upon return. During the actual combat and mass killing,
there was not much opportunity for organized socialization, such as the
diaspora had experienced. Rwanda was in a state of uncertainty and
chaos, with more than 20 percent of the population displaced. Rather,
the period from 1990 to 1994 highlights the results of the transformative
learning that occurred in exile. As the war progressed, the Tutsi rebel
army, led by Paul Kagame, demonstrated the value of the exile experience
in terms of military cohesion and capability. It appears that Kagame also
used this time to plan a strategy for implementing his vision once in
power. An essential component of this vision included the reintegration
of returned refugees as significant political actors (Kinzer 2008).
Refugee militancy in exile and upon return
113
Transnational interactions among mechanisms
The Hutu refugees in Zaire also attempted to transform their institutions to promote militarized return. The militants’ goal was to return
in military victory to Rwanda and complete the genocide against the
Tutsi. To that end, the militant leaders transferred the government
structure from Rwanda and established a government in exile. The
interaction of the socialization, resource distribution, and security
entrapment mechanisms reinforced the power of the quasigovernment. The use of propaganda transformed refugee world views
by describing the genocide in terms of a civil war. Leaders supplemented this propaganda with manipulation of humanitarian aid
resources to benefit supporters. Refugees were torn between the security entrapment created by militant coercion in the camps and the
security fears of what awaited them in Rwanda. However, pressure
such as forced recruitment and extortion reduced the utility of other
socialization programs (Boutroue 1998).
The transnational nature of this conflict became more pronounced
as the Hutu militants increased cross-border attacks. In 1996, the
RPF perceived a significant existential threat from the hostile Hutu
state-in-exile in Zaire. In alliance with Zairian insurgents, the RPF
invaded Zaire and forcibly repatriated hundreds of thousands of
refugees back to Rwanda (UNHCR 2010b). In the coercive
repatriation of the Hutu refugees, one sees the unusual phenomenon
of a cross-border conflict between two state-in-exile refugee groups.
Andrea Purdeková explains: “Rwanda represents a unique case
study not only due to the scale and complexity of movement in a
relatively brief period, but also because the victors of the war were
themselves returning exiles” (2008, 5). Mamdani concurs that
“what has come to be known as the RPF invasion of Rwanda also
needs to be understood as an armed repatriation of Banyarwanda
refugees from Uganda to Rwanda” (Mamdani 2001, 160). The
return of militarized Tutsi refugees in Rwanda demonstrates
the significance of socialization and resource distribution, within a
context of insecurity. The RPF, as a mobilized diaspora in Uganda,
developed the political, military, and social organization that underpinned its successful invasion. Upon their return, the coercive Hutu
militants lost their ability to control the refugee population, solidifying the power of the RPF.
114
Kristian Berg Harpviken & Sarah Kenyon Lischer
Reintegration and consolidation of power, 1994–2010
Socialization
Upon taking power, the RPF government clearly realized the impact of
socialization in exile, likely due to its own experience in Uganda.
A report on post-genocide violence commissioned by the Rwandan
government proposed a wide-ranging socialization campaign to reduce
what it described as “wickedness.” Recommendations included:
The education and the sensitization of people through all agents and organs
of socialization such as solidarity camps (ingando) . . . sketches and plays,
movies, profane songs, religious songs, media, local and national festivals,
civil society (religious confessions, local and international NGO’s, various
associations including those of national unity and reconciliation), operational political parties on the ground, public administration, public and
private firms and education system (secondary schools, higher institutes
and universities). (Republic of Rwanda. National Unity and Reconciliation
Commission [NURC] 2008, 114)
An observable implication of the socialization mechanism includes
public and private discourse which reflects changed world views. In
Rwanda, group labels radically changed to reflect the government’s
attempt at transformative learning in the realm of ethnic identity
(typified by the quote above). Following the genocide: “The state
language in Rwanda . . . divide[d] the population into five categories:
returnees, refugees, victims, survivors, and perpetrators” (Mamdani
2001, 266). Within the unspoken designation of Hutu and Tutsi, the
categories played a significant role in the political landscape. For
example, the term “survivor” described a Tutsi civilian who lived in
Rwanda during the genocide. Mamdani comments that no Hutu are
labeled survivors since “the assumption is that every Hutu who
opposed the genocide was killed . . . The dilemma is that to be a Hutu
in contemporary Rwanda is to be presumed a perpetrator” (Mamdani
2001, 267; italics in original). Subcategories among the Hutu also
often referenced a person’s judicial status. The term “perpetrator”
(or, more colloquially, “génocidaire”) referred to those convicted
(or even suspected) of participation in the genocide.4 Surprisingly, the
4
As pointed out by Lee Ann Fujii, those labels were not necessarily accurate since
some people were imprisoned on false charges and many génocidaires had
never been imprisoned (personal communication, 2010).
Refugee militancy in exile and upon return
115
categories of “refugee” and “returnee” held less resonance for Rwandan
citizens than might be supposed, considering the extensive usage of the
terms in secondary literature. Perhaps that was because such a large
percentage of the population had an experience of exile.5
More formal discourse included laws against so-called “divisionist”
speech, which limited public discussions of ethnicity, and constrained
private communications as well.6 These laws provided for severe
penalties of up to 25 years in prison for offenses ranging from murder
to “threatening, intimidating, degrading through diffamatory [sic]
speeches, documents or actions which aim at propounding wickedness
or inciting hatred” to “marginalizing, laughing at one’s misfortune,
defaming, mocking, boasting.” Remarkably, children under 12 years
old could be convicted of the crime of genocide ideology and taken to a
“rehabilitation centre” for up to one year (Republic of Rwanda 2008).
Those many cleavages and unspoken divisions were products of
and affected the process of socialization. This is apparent in the
extent to which these guidelines were followed in public and private
conversations. The observable implications of this discourse are of a
socialization that relies on coercion in an attempt to change popular
conceptions of ethnicity. It is difficult, however, to ascertain the extent
of such transformations (see also Bakke, this volume); in a context of
insecurity, this is made more difficult since informants do not speak
freely. Outwardly, the socialization has succeeded in transforming
discourse. It will take a longer time to see if such measures can create
transformative learning among a reluctant, suspicious population.
The new institutions which developed reflect the operation of the
socialization mechanism. The centerpiece of the anti-genocide ideology
campaign was the government’s rigorous program of returnee socialization called ingando which aimed to deprogram the Hutu extremist
socialization that occurred in exile. These non-voluntary re-education
camps included “sensitization training” covering “Rwandan history, civic
education, national unity and reconciliation, gacaca, micro-financing
and public health (particularly HIV/AIDS)” (Waldorf 2009, 11).
Purdeková describes these camps as “a space of tightly upheld
military-like discipline and instruction rather than debate” (2008, 18).
5
6
Interviews, Rwanda, April May 2009.
These restrictions on free speech meant that most of my informants agreed to
speak only on condition of confidentiality.
116
Kristian Berg Harpviken & Sarah Kenyon Lischer
Repatriating ex-combatants spent two months in a separate ingando
camp upon their return to Rwanda. According to MONUC, the
UN peacekeeping force in the Democratic Republic of Congo (DRC,
formerly Zaire), the program effectively prevented violence after
return.7 In addition to former soldiers and militants, the ingando
program expanded to target prisoners in transition between partial
and full release, as well as civilians such as teachers, civil servants, and
post-secondary students.
In judging the success of the socialization, one can observe outcomes
such as the level of violence and nature of public discourse. The
ultimate goal of the ingando program, to change the thinking of
the participants, is not easily observable. Interviews with ingando
participants and with experienced observers also provide insight into
the process (Thomson 2011). In the short term, violence involving
repatriates is low, although a multitude of factors (including coercion)
can contribute to that.
Resource distribution
In some instances, resource distribution works through providing
patronage jobs, yet in others it operates by cutting off resources selectively. Excluding or withholding resources can influence the potential
for remilitarization. An indicator of this mechanism at work is the
selective distribution of resources to potential enemies in exchange
for pacification (if not loyalty).
The Rwandan government, through the disarmament, demobilization,
and reintegration (DDR) program, was the primary provider of resources
to former combatants. After completing the ingando training, the
ex-combatants received basic needs kits. Some received further funding,
depending on their rank and affiliation (Waldorf 2009, 10–11). Using
reintegration packages and other demobilization resources, partly funded
by the World Bank, the government fostered dependency and encouraged
compliance from the ex-combatants. Mgbako comments that “the
government is offering a chance at reintegration and economic aid to all
rebels, fearing that if it does not assist these rebels in peaceful resettlement
they will revitalize the insurgency” (2005, 210).
The second main resource needed by returnees in Rwanda was land.
Rwanda is the most densely populated country in Africa and the vast
7
Interviews, UN officials, Kigali, Rwanda, April 2009.
Refugee militancy in exile and upon return
117
majority of people rely on subsistence farming. Land issues were
immensely complicated by the numerous migration flows in and out
of Rwanda since independence. In terms of resource distribution, the
government took control of this valuable resource, in part due to weak
private property laws.
Following the 1990–1994 civil war and genocide, Rwandan Tutsi
who had lived in exile for up to 30 years claimed the right to the lands
they abandoned, even though they were now occupied. Minear and
Kent explain that:
the old caseload refugees posed a very complex problem . . . these returnees
after so many years in exile had high expectations. Those among the 600,000
who lacked housing, employment, and land – or whose homes and lands had
been occupied in the interim – represented a potentially explosive political
and emotional issue. (1998, 64)
As discussed above, the ability to distribute resources increases the
power of a militarized group. The RPF needed to follow through on
its promises to retain the loyalty of its followers. The Arusha Peace
Agreement of 1993 outlined a power-sharing deal that attempted to
defuse that thorny issue. The Accords mandated the following:
All refugees shall therefore have the right to repossess their property on
return. The two parties recommend, however, that in order to promote social
harmony and national reconciliation, refugees who left the country more
than 10 years ago should not reclaim their properties, which might have
been occupied by other people. The Government shall compensate them by
putting land at their disposal and shall help them to resettle. (Arusha Peace
Agreement 1993, Chapter I, 1.1.4)
In the context of extreme land scarcity, massive migration flows, and a
weak transitional government, the vision of the Agreement was not
implemented.
The shortage of land and perceived injustice of the resource distribution continued to threaten political stability in Rwanda in the years
following the genocide. One can observe the changes wrought by the
preferential distribution of resources in the wide-scale bulldozing of
slums in Kigali and relocation of mostly Hutu residents to distant,
inhospitable areas. The cleared areas are then awarded to government
supporters for so-called economic development projects. An Anglican
priest, and genocide survivor, remarked caustically that soon the entire
118
Kristian Berg Harpviken & Sarah Kenyon Lischer
map of Kigali would read “no poor people allowed.”8 UNHCR
warned that “the threat to stability of the reintegration process is
real by creating categories within the returnee community (between
those having received a house and those who haven’t)” (UNHCR
2002, viii). A 2007 Norwegian Refugee Council report worried
that the complicated relationship between ethnicity and land issues is
“disregarded and even suppressed by the government” (Bruce 2007).
To the extent that the resource distribution patterns reinforce cleavages
among competing groups, former militants might be induced to take
up arms to rectify perceived injustices.
In Rwanda, socialization was the foremost mechanism influencing
returnee violence, but the role of resource distribution was essential in
enabling leaders to capitalize on the socialization process. As the Tutsi
refugees reached consensus on the need for military return, the RPF leaders
were able to provide material benefits (in particular the promise of future
benefits) as incentives to their followers. When they invaded Rwanda,
the returnees were able to take over the land of the Hutu civilians displaced
by the fighting. As part of the 1993 Arusha Peace Agreement, the RPF
negotiated for restitution of property or government compensation for
the Tutsi refugees who had been displaced to Uganda. Thus the returnees
were able to mitigate fears of security entrapment by remaining within
the group formed during the socialization process in exile.
Concluding remarks
This chapter examined a form of conflictual transnationalism: the
potential for violent engagement upon return from militarized exile
contexts to the country of origin. Our approach privileges the role of
mechanisms in understanding returnee violence. The socialization,
resource distribution, and security entrapment mechanisms help elucidate the conditions under which previously militarized refugees engage
in violence upon return in both Afghanistan and Rwanda. At the
empirical level, our findings highlight the advantages of a mechanism
approach, particularly when dealing with complex levels of variation
and interaction within and between cases.
We conceive of socialization as transformative learning and are
particularly interested in socialization that encourages militant attitudes.
8
Interview with the author, Kigali, Rwanda, May 7, 2009.
Refugee militancy in exile and upon return
119
This occurs most readily in the seclusion and uncertainty experienced
by many refugee populations. Such socialization can lead to violence
upon return to the home country. Our second mechanism focuses on
the way in which the distribution of resources – widely defined to
include anything from emergency aid, to jobs and property – stimulates
recruitment and allegiance, building interdependence between refugees
and the leaders of militant groups. The final mechanism discussed is
security entrapment, which plays out when return exacerbates external
threats to individuals with the militant group being the only possible
guarantor. Security entrapment can also create incentives for militarized return when refugees perceive threats while in exile.
It is unlikely that we will find trajectories of refugee and return
militancy where only one of the mechanisms is operating. There is a
logical sequence in which socialization takes place mainly prior to
flight and in exile, resource distribution happens in exile and (in part)
upon return, whereas security entrapment happens mainly after return.
But the mechanisms can also interact jointly, as when Afghan refugees
in Pakistan in the 1980s were enrolled in socialization campaigns as a
reflection of controlled resource distribution, or when Rwandan Hutus
in Zaire were affected by all three mechanisms (nearly) simultaneously.
In both our cases, militarization hinged on a combination of socialization and resource distribution; security entrapment, while often
important, clearly played a secondary role.
In exploring the mechanisms by which mobilized refugees contribute
to violence upon return, the chapter contributes to larger debates in at
least two substantive bodies of literature: (forced) migration as well
as civil war and transnationalism. To the former, we contribute an
in-depth framework for exploring political violence involving militant
refugees upon their return. This is an extension of the emerging realization that refugees have agency – and that such agency can sometimes
have a “dark side.” To the debate on civil war and transnationalism,
our findings highlight the dynamic nature of the transnationalism that
feeds into conflict. We do so by examining the mechanisms that foster
militancy throughout the whole refugee cycle (from pre-flight via exile
to post-return), as well as the ways in which the relations between exiles
and others (whether friendly or hostile) affect violence. The systematic
application of a mechanisms-based framework, while empirically
demanding, does give us a better understanding of the dynamism of
the processes that yield political violence.
|
5
Rebels without a cause? Transnational
diffusion and the Lord’s Resistance
Army, 1986–2011
hans peter schmitz
In October 2005, the International Criminal Court (ICC) unsealed
arrest warrants for five leaders of the Lord’s Resistance Army (LRA).
From the late 1980s and early 1990s onward, the LRA committed
widespread atrocities against civilian populations in the border regions
of Sudan and Uganda. During the past decade, the rebel group –
estimated to number a few hundred to possibly several thousand
fighters – committed similar acts of violence in the Democratic Republic of Congo (DRC) and most recently in the Central African Republic
(CAR). Led by Joseph Kony, the LRA has evaded several high-profile
military operations by the Ugandan, Sudanese, and Congolese military
and has participated in several peace negotiation efforts without ever
agreeing to a settlement ending the bloodshed. Thirty years of violence
have resulted in an estimated 100,000 deaths, more than 30,000
children abducted, and many more displaced, often for decades forced
into refugee camps (Okiror 2007).
While a long list of celebrities and state leaders have taken up the
cause of the victims turned into child soldiers and sex slaves, the
international community is not getting closer to ending this violence.
The military has pushed the LRA out of Northern Uganda (in the early
1990s), out of Southern Sudan (in the early 2000s), and now out of the
DRC (around 2008), but this response has been ineffective in protecting scores of civilians in these border regions from atrocities. At the
same time, non-military efforts such as the ICC indictments and subsequent direct peace negotiations between the LRA and the Ugandan
government from 2006 to 2008 have fallen short of creating the right
incentives for LRA leaders to lay down their arms. What explains this
particular pattern of violence, in particular its persistence as well as
changes in the use of certain strategies, including committing atrocities
against civilians and abducting children and adults?
To understand better new forms of post-Cold War violent conflict, a
growing body of research focuses on the transnational dimensions of
120
Transnational diffusion & the Lord’s Resistance Army
121
these struggles. Many of the recent quantitative studies addressing such
questions have highlighted the role of diaspora support or the ability of
rebel groups to retreat across state borders. But most of these studies
are content with claims that such factors matter for outbreak and
perpetuation of violence without showing how they specifically play
into the mobilization of resources, changes in framing the violence, and
choices of targets and strategies (see Checkel, this volume). Such analyses also largely fail to look at the interactions between transnational
mechanisms and other, more locally driven factors that may mediate or
even render external influences ineffective. Finally, rarely do studies
explore the interaction between transnational mechanisms sustaining
the violence and countervailing transnational efforts designed to end it.
The LRA is a compelling and “hard” case for understanding the
effects of transnational influences because its leadership deliberately
limits any exposure to external influences. While it is impossible to
assess the extent of many transnational ties that may have shaped the
actions of the LRA (see also Bakke, this volume), Joseph Kony consciously avoided sustained and complex interactions with the outside
world, even during the 1990s when the rebel group relied heavily on
military support by the Sudanese government. Existing research on the
LRA has frequently been preoccupied with its extraordinary violence
against civilians as well as the perceived lack of a rational explanation
for its goals and strategies. This type of scholarship views the LRA as a
post-Cold War movement that no longer has a clear political agenda
such as attaining some form of independent rule or seizing national
power. However, more recent ethnographic studies of the LRA reject
such claims of irrationality and have provided more contextual and
historical accounts for its actions.
The present chapter adds to this reassessment a specific focus on
transnational mechanisms and their impact on shifts in the levels of
violence, the LRA leadership’s framing of the struggle, and choices
regarding targets and strategies. While greater attention to local context has helped us better understand the ineffective nature of international responses (Autesserre 2009), an added focus on transnational
mechanisms sheds light on how shifts in a rebel movement’s environment are equally relevant to developing more effective interventions.1
The continued international military response reflects a limited grasp
1
Thanks to Matt Evangelista for discussions on this point.
122
Hans Peter Schmitz
of both the local and the transnational dimensions of this conflict. As a
result, the international community time and again fails to develop
effective strategies to protect civilians. President Obama’s October
2011 decision to deploy an additional 100 US armed forces personnel
to eliminate the LRA (Obama 2011) represents a continuation of such
failed external efforts.
I proceed in the following manner: the next section summarizes
specific shifts in the LRA’s behavior and how transnational mechanisms provide important insights into those changes. The three empirical
sections that follow are organized around the topics of framing and the
justification of violence, strategies used, and resource mobilization.
The first empirical part focuses on how transnational mechanisms
affected the initial framing of violence; the second explains how shifts
in the use of violent means are a result of emulation and social
adaptation, each of which operationalizes the broader category of
transnational diffusion (Checkel, this volume); the third looks at shifts
in resource mobilization over time, especially the relational diffusion of
material and immaterial resources mobilized by external supporters
and opponents.
The fourth section returns to the topic of framing and explains why
the indictments of the ICC in 2005 and the peace negotiations from
2006 to 2008 have failed to adequately address the violence. The
dominance of an international approach devoid of efforts to build
meaningful transnational relations with those seeking to end the violence at the local level continues to give the advantage to the LRA and
its efforts to exploit transnational opportunities for its own goals.
Transnational mechanisms and non-state violence
Research on social movements has focused particular attention on
three aspects of domestic activism that may be shaped by outsiders
through transnational mechanisms (see also the chapters by Adamson
and Bakke in this volume). This chapter examines competing explanations accounting for: (1) shifts in the goals; (2) the framing, targets,
and collective means used; and (3) the mobilization of resources. It not
only investigates how external supporters of a rebel movement can
shape such shifts, but also includes the often ambiguous effects
of international efforts to end the violence, in particular, through
global human rights mobilization. The analysis provides a broader
Transnational diffusion & the Lord’s Resistance Army
123
understanding of how different and often uncoordinated international
actions interact over time – including early missionary teachings in the
region, the activism of the Acholi diaspora community, the emulation
and learning of tactics and strategies from other rebel movements in
neighboring nations, the impact of transnational human rights mobilization, and the practice of East African governments to use rebel
groups to engage in forms “proxy warfare.”
Tracing the process of how transnational mechanisms shape the
goals, strategies, and resources of rebel movements creates challenges
with regard to the availability and reliability of data.2 Thanks to the
work of local and international relief agencies, information about
shifts in the targeting and strategies of the LRA is relatively reliable
since episodes of abductions and violence are well documented. The
main difficulty in assessing shifts in framing is over the authenticity of
any written or spoken statements made on behalf of the LRA. Similar
challenges arise when attempting to assess the amount of resources the
LRA received from external sources at any given point in time.
The chapter’s empirical evidence shows that transnational mechanisms play an important, but not necessarily overwhelming role in
shaping behavior and outcomes. They are mitigated by pre-existing
local conditions and their significance varies not only across the issues
of framing, resources, and strategies, but also over time. Following the
Nome and Weidmann and Bakke chapters (this volume), I argue that
processes of relational diffusion – that is, the transfer of information or
resources through personal networks and social bonds – are more
robust. Indeed, efforts relying on targeted, non-relational diffusion –
where people in one locale learn from people elsewhere through television, the radio, newspapers, and the internet – are not only less
effective, but also often counterproductive. While diaspora mobilization was more prominent during parts of the conflict, its effects were
much less important in this case than in Adamson’s study of the
conflict between the PKK and the Turkish government (Chapter 3, this
volume). Table 5.1 summarizes the core claims about the relative
importance of transnational mechanisms in accounting for significant
shifts in the key activities of the LRA with regard to goals, strategies,
and resource mobilization.
2
On how to define and observe mechanisms, see the introductory chapter by
Checkel.
Table 5.1 Transnational mechanisms and shifts in framing, resources, and repertoires
Observable shift to be explained
Transnational mechanisms at work
Competing accounts
Framing and
justification of
violence
(section 1)
1987 1995: Development of a loosely Christian
ideology and “spiritualist” movement
1995 2003: LRA manifestos emphasize
secularism and political goals
Relational diffusion
Incoherent “cult;” “rebels
without a cause”
Manifestos are not authored
by LRA leadership/
represent mere “window
dressing”
Repertoires of
violence
(section 2)
Early 1990s: Increased use of atrocities against
civilians, including own Acholi population
Late 1990s: More limited use of atrocities, but
continued abductions
Resource
mobilization
(section 3)
1990s 2002: LRA moves to camps in Southern
Self sufficient strategy based
Relational diffusion to supply small arms by
Sudanese government and/or private dealers;
Sudan and gains military support by Sudanese
on abductions and looting
government
lack of government control over border regions
2002 2005: LRA loses the material support from
Relational diffusion to supply small arms by private Self sufficient strategy based
the Sudanese government; moves to camps in the
dealers; lack of government control over border
on abductions and looting;
DRC
regions
Amnesty offers deplete force
Lack of relational diffusion and international
2005 today: LRA participates in peace
Military pressure; LRA faces
negotiations mediated by Southern Sudanese
capacity to address grievances; non relational
defeat
government; LRA moves to CAR after failed
diffusion dominates international approach
talks end in 2008
focused on prosecution, not protection
Frames of global 2003 today: Increasing dissent within LRA
justice and
following amnesty offer and ICC indictments
peace (section 4)
Non relational diffusion (“shaming” by human
rights groups) and relational diffusion by
diaspora
Emulation and situational learning in the context of
the civil war in Southern Sudan
Non relational diffusion (“shaming” by human
rights groups)
Irrational behavior of a
“cult” and its leader
Material rewards for not
targeting civilians
Non relational diffusion (threat of prosecution by
Material incentives and
ICC); relational diffusion during peace negotiations
concessions in the context of
negotiations; military
pressure
Transnational diffusion & the Lord’s Resistance Army
125
Framing and justification of violence, 1987–2003: LRA
manifestos and the transnational dimensions of the uprising
This section provides evidence on the role of transnational mechanisms
in shaping two shifts in the framing of LRA violence. I first describe how
taking up arms against Museveni’s new government as well as the LRA’s
own Acholi population was rationalized in the late 1980s; second,
I discuss how increased diaspora and international human rights activism caused a more defensive framing of violence in the mid to late 1990s.
A third shift in framing following the ICC indictment of five LRA
leaders in 2005 is discussed in the final empirical section.
Many external observers have been drawn to this conflict because of
the extent of atrocities committed as well as the difficulties of establishing clear motives and justifications. Unlike the Chechnya case
(Bakke, this volume), religion as a transnational force plays a more
subtle role in the framing efforts of the LRA, and a number of external
forces – in particular, governments of the region engaged in a “proxy
war” (Prunier 2004) – shaped both resource availability and framing.
While a majority of journalistic as well as academic coverage of the
LRA tends to focus exclusively on the atrocities, this section will
highlight in what ways attention to a combination of local grievances
and transnational mechanisms enables an understanding of the significant shifts in the LRA’s behavior and framing of the violence over time.
The main obstacle to tracing shifts in frames used by rebel groups
such as the LRA is the limited access to documents such as manifestos
or other forms of self-representation. Many websites posting information about the LRA were shut down after September 11, 2001 when
the US government added the rebel group to its list of terrorist organizations. The dearth of first-hand information has led to significant
distortions in the “official discourse” (Finnström 2008, 99) prevalent
in the international media and among scholars claiming that public
statements are useless and the LRA leadership is irrational. The media
regularly portrays the LRA “as a barbaric and insane cult” (Allen
2006, 25) and even more experienced researchers claim that “the
LRA has no political program or ideology, at least none that the local
population has heard or can understand” (Gersony 1997, 59).3
3
This view continues to dominate the current representation of the LRA by the
Enough Project, a leading transnational activist group with the goal of “ending
126
Hans Peter Schmitz
More recently, a number of scholars have defied this mainstream
view (Bevan 2007; Allen and Vlassenroot 2010) and offered more
detailed analyses of LRA manifestos and shifts in how its leadership
framed its violent struggle. These challenges to the “myth of madness”
do not excuse LRA atrocities, but find current accounts unsatisfactory
because “we need to pay attention to the power relations and structural circumstances that promote such persons’ positions” (Finnström
2008, 129). While the authenticity of many LRA statements is contested, scholars with extensive field experience maintain they can be
traced back to its leadership; moreover, the origin of the manifestos is
less important than how they were used (or ignored) by local and
global audiences (Finnström 2008, 120).
The dominant narrative that emerged in the mid 1990s was driven
primarily by major human rights organizations, the Ugandan government, and international media, and focused almost solely on the atrocities and abductions of minors (Amnesty International 1997; Human
Rights Watch/Africa 1997). Dolan notes that Amnesty International
corrected its initial bias in a follow-up report two years later (Amnesty
International 1999) “concentrating on government violations” (Dolan
2009, 229), but most of the attention by Northern publics remained
focused on the LRA and its leader Joseph Kony.4 Moral outrage about
the welfare of children proved to be an effective tool in mobilizing
Western audiences, but crowded out other important local grievances
and interests. Western audiences found comfort in their preconceived
notions about the continent, concluding that “the conflict is bizarre,
but Africa is simply like that” (Ehrenreich 1998, 84). Aligned with the
political interests of the Ugandan government, the aid community
active in Northern Uganda during the 1990s followed an “inverse
relationship between . . . organizations’ level of operations, scale and
capacities in terms of funding and staffing, and their willingness to
speak up about those aspects . . . which involved criticism of the
Government of Uganda” (Dolan 2009, 228). Local NGOs were most
critical, national ones less so, and international NGOs as well as UN
agencies almost entirely “laid the blame for all suffering at the feet of
4
genocide and crimes against humanity.” On its website, it claims: “The LRA has
no clear political agenda.” See www.enoughproject.org/LRA (accessed July 31,
2010).
It took Human Rights Watch until 2005 to publish a more comprehensive report
on the violence in Northern Uganda (Human Rights Watch 2005).
Transnational diffusion & the Lord’s Resistance Army
127
the LRA” (Dolan 2009, 229). The idea of “rebels without a cause”
resonated with Northern publics’ perceptions of Africa and served the
political and economic interests of the Ugandan government and many
NGO and human rights activists.
Following the military victory of the NRM/A, local populations in
the northern and north-western parts of the country quickly became
concerned about their diminished role in national politics and the lack
of development compared to the prospering southern parts (OmaraOtunnu 1992). The NRM government ended the long-standing northern domination (Acholi and Langi) of the Ugandan army and state
(Omara-Otunnu 1987) and southern elites now took control of
national politics. After decades of violence and civil war under the
dictatorial regimes of Idi Amin (1971–1979) and Milton Obote (1981–
1985), the NRM/A under the leadership of Yoweri Museveni
decisively defeated government forces in late 1985 and gained control
of the capital, Kampala. Significant insurgencies organized by military
personnel of prior regimes persisted in many parts of the country until
the early 1990s. Joseph Kony and his rebel movement differed from
other violent rebellions during this time by combining local grievances
about the marginalization of Northern Uganda with the use of elaborate religious rituals designed to challenge the traditional dominance of
local Acholi chiefs and convince individual fighters that they were
invincible (Behrend 1999; Cline 2003). When a rebel group called
Holy Spirit Movement (HSM) led by Alice Auma (Lakwena) was
beaten by NRA forces in 1988, Kony became the only remaining rebel
leader capable of organizing violent resistance. Until the mid 2000s,
the LRA regularly attacked Northern Uganda and close to 2 million
Ugandans were forced into what the Ugandan government “euphemistically calls . . . ‘protected villages’” (Branch 2007, 181). This persistent state of insecurity left Northern Uganda economically devastated
which regularly pushed young males into the arms of rebels and
created more violence and insecurity.
Framing the uprising: missionary teachings, marginalization,
and the crisis of traditional societies
The marginalization of Northern Uganda, going back to the colonial
period (Mamdani 1984), provided the initial and most enduring frame
justifying the violent struggle against the Museveni government in
128
Hans Peter Schmitz
Kampala since the mid to late 1980s. British colonial rule gave preferential treatment to the Southern Baganda and the “the northern region
served mainly as a reservoir for cheap labor, from which Britain
recruited its soldiers” (Van Acker 2004, 341). After independence, this
imbalance led to the establishment of consecutive military regimes
under Idi Amin and Milton Obote solidifying the idea of “competitive
retaliation on an ethnic basis” (Van Acker 2004, 340).
Successive leaders of rebel movements in the North frequently gave
meaning to the long-standing grievances by adding mystical and religious overtones to support their peculiar claims to leadership. Many
observers have used the frequent biblical references as a reason to
dismiss them as fundamentalist (Adam et al. 2007) or “millenarian”
(Chabal and Daloz 1999, 86). Yet, the logic of this frame actually
“echoes more than one century of missionary teaching” (Finnström
2008, 108) and confirms the importance of transnational processes of
relational diffusion that show in what ways “this kind of terror . . .
might prove to be linked to the logic of the world system and, as such,
to be less exceptional than hoped” (Doom and Vlassenroot 1999, 7).
While the grievances of the Acholi people after the NRM/A victory
were primarily local and national, the framing of the particular
response developed by the LRA leadership was derived from a global
structure of values that had been diffused into the Ugandan polity for a
century using mechanisms of socialization, reward, and severe punishment. Many aspiring new leaders emerged in the North during that
time and shared similar visions as those expressed by Kony and the
LRA. For example, “Alice Lakwena’s Holy Spirit Movement was the
outcome of elements of missionary Christianity interacting with indigenous cosmology” (Van Acker 2004, 346).
Instead of understanding the unique combination of local grievances
and transnational religious ideas, Western perceptions were primarily
informed by pictures of the abducted and maimed. Finnström details
how popular books and even feature films about the conflict perpetuated the contrast between a “dark continent” and the help provided by
non-Africans, including the popular story of European nuns searching
for 30 girls abducted from the Aboke missionary school (de Temmermann 2001; Finnström 2008, 110). The point here is not to question
the good intentions of external actors willing to act, but to elaborate
how ignoring the larger, transnational context reproduced problematic
assumptions about the conflict. President Museveni and his
Transnational diffusion & the Lord’s Resistance Army
129
government promoted this particular perspective of Africa, portraying
his regime as modern and rational and declaring that his enemies in the
North were “intoxicating poor peasants with mysticism and incredible
lies” (Museveni 1992, 115). Internally, Museveni used the violence in
the North to consolidate his power base by presenting himself as the
leader guaranteeing peace (for the South) and protector “from the
Acholi and other wild northerners” (Allen 2006, 48). Museveni skillfully reinforced Western perceptions of an unresolvable conflict while
also falling in line with US-driven global security interests focused on
military solutions to terrorist threats (Schmitz 2006).
The most promising peace negotiations between the Ugandan government and the LRA during this time period took place in 1994, when
the Acholi Betty Bigombe, a minister of state in Museveni’s government appointed in 1988 to end the rebellion in the North, met directly
with Kony on several occasions. Against much male prejudice,
Bigombe built crucial trust among the Acholi prior to initiating the
peace talks, which followed shortly after “Operation North,” the first
major military action by the Ugandan army, failed to defeat the rebels
in 1992 (O’Kadameri 2002, 35). In a response to a letter sent by
Bigombe, Kony indicated that he would need to “receive guidance by
the Holy Spirit” and by October 1993 pre-negotiations on security
questions took place. During a meeting with Bigombe and Acholi
representatives in January 1994 in Gulu, Kony blamed the Acholi
community and its elders for the violence because they abandoned
the LRA (O’Kadameri 2002, 40). The talks soon deteriorated, with
LRA leaders accusing government officials of humiliating them and the
government suspecting the LRA of establishing contacts with the
Sudanese government. Museveni visited Gulu in February 1994
announcing that the army would defeat the LRA if it did not lay down
its arms within seven days.
Defensive framing: early human rights mobilization
and diaspora activism
Following the breakdown of the first peace negotiations and the subsequent increase in violence, the diaspora community in collaboration
with Acholi civil society groups initiated in 1996 the “Kacoke Madit”
(big meeting) process, intended to champion peace and reconciliation.
As the global human rights community began to take note of the
130
Hans Peter Schmitz
atrocities and began to publicize the systematic attacks on civilians, a
second shift in framing the violence took place. The LRA now supplemented its demands for greater political participation with a denial of
abuses and a rhetorical emphasis on universal human rights. For the
first time, the conflict became the subject of widespread discussion
beyond the region and diaspora community.
At the first Kacoke Madit meeting held at the University of London,5
James Obita claimed to represent the LRA/M and defended the refusal
to surrender, arguing that those being offered amnesty in the past were
killed by Museveni’s troops (Obita 1997). Obita portrayed the emergence of the LRA as a rightful act of self-defense against the aggression
of the Ugandan military, although his use of the term “NRA-Tutsi
army” exposed his own ethnically biased discourse. The speech listed
several political demands, including the immediate restoration of multiparty democracy, Museveni’s resignation, formation of a government
of national reconciliation, ending gross human rights violations, and
economic prosperity for all Ugandans. While he acknowledged the
suffering of Acholi civilians, the speech only admitted to “instances
of indiscipline” on the part of the LRA while lamenting a failure to
adequately investigate atrocities committed by government troops.
Obita claimed that the LRA leadership took harsh measures against
anyone in the organization who may have “reportedly taken people
against their will or committed atrocities. Disciplinary measures,
including even execution by firing squad of LRA fighters who have
grossly abused civilians have taken place in our camps.” Obita also
claimed that investigations of atrocities were currently under way, but
there is no subsequent record about their outcomes. With regard to
mutilations, Obita emphasized that “no official policy of carrying out
such atrocious acts” exists and he blamed government soldiers “masquerading as LRA/M” for several such incidents reported in 1996.
In 1997, a LRA manifesto demanded multiparty politics and constitutional federalism, expressed support for human rights and anticorruption, and demanded a separation of the military from the judiciary and the executive (“From LRA/M to all Ugandans,” cited in
Finnström 2008, 122). Recalling how previous rebel movements
5
Two subsequent meetings in 1998 (London) and 2000 (Nairobi) were attended
by lower ranking representatives of the Ugandan government, Acholi community
members, and civil society groups. See www.km net.org.uk (accessed November
20, 2010).
Transnational diffusion & the Lord’s Resistance Army
131
disappeared without accomplishing their goals, a 1999 statement –
responding to an amnesty offer – insisted that “[w]e are not going to
lay down our arms as long as Museveni is still in Uganda as president”
(LRA letter, cited in Finnström 2008, 120). During this time, the LRA
not only committed the majority of its atrocities in Northern Uganda,
but simultaneously released the most detailed manifestos with a distinct new framing based on an emphasis on secularism and global
values. The blatant gap between the violence against the civilian population and the claims in the manifestos widened as the written statements increasingly responded to criticisms of LRA atrocities and added
explicit denials of the official discourse about their alleged motives.
Finnström cites an 18-page political manifesto circulated in late 1999
which criticized the Ugandan army’s recent military actions in the
DRC, demanded multiparty elections, and claimed that structural
adjustment programs had the most adverse effects on local populations
in the periphery of Uganda (Finnström 2008, 125–126). The manifesto
also outlined “brief descriptions of LRA/M’s economic programs and
proposed policies on education, agriculture, health, land and natural
resources, infrastructure, commerce and industry, and defense”
(Finnström 2008, 123). Other issues raised regularly included demands
for national unity by promoting inter-tribal marriages and language
instruction, education for all, an independent judiciary, the formation
of an ethnically balanced national army, and policies to “end the use of
witchcraft and sorcery by promotion of the Ten Commandments”
(cited in Allen 2006, 43).
Along with promoting a broader political agenda largely ignored by
the Ugandan government and external observers, the manifestos also
explicitly denied that the LRA was motivated by fundamentalist Christian ideology. Finnström cites now defunct websites accessed in 1999
as well as written materials circulated under Kony’s name stating that
“I would like to strongly deny that these members [of the LRA, HPS]
are or in any way have the intentions of becoming Christian Fundamentalists” (Finnström 2008, 124). At the height of the propaganda
campaign in early 1999, the LRA broadcast daily for one hour on a
station called Radio Free Uganda until the signal was blocked by the
government. The program “belies the claim that the LRA lacked any
political position” (Dolan 2009, 85) and consisted primarily of political attacks against the government. Topics discussed included excessive military expenditures, lack of democratic accountability, and the
132
Hans Peter Schmitz
encouragement of mob justice by government officials (Dolan 2009,
84–85). The program was also used to refute claims circulating among
the local population that the LRA was involved in slave trade and
resorted to cannibalism. Most importantly, the broadcast denied
claims by the government that the LRA was on the brink of military
defeat. In October 2002, the Ugandan government prohibited another
radio station in Gulu from airing any LRA statements after several callins by Joseph Kony and other rebel leaders during live broadcasts.
The LRA’s intensified public relations efforts in the mid to late
1990s coincided not only with greater exposure of the atrocities in
Western media, but also with increased military pressure by Uganda
forces as well as greater international efforts to put an end to the
violence. By 1999, Northern Uganda had experienced almost one year
of peace as the Ugandan army was able to push the LRA deep into
Southern Sudan. In December, the Carter Center facilitated a peace
accord between Sudan and Uganda (Neu 2002), although the preceding negotiations excluded representatives from the LRA or the Sudan
People’s Liberation Army (SPLA). In January 2000, the Ugandan
parliament adopted the Amnesty Act, offering a blanket amnesty to
all rebels willing to lay down their weapons. While the top leadership
of the LRA rejected the offer, many in the second or third tier left the
rebel group, often citing the amnesty as the most important factor for
their decision to return to civilian life (Conciliation Resources 2006).
Dolan reports that the number of LRA members taking advantage of
the offer increased from 1,086 in 2002 to 3,601 in 2004 before
dropping to less than 100 in 2007 (Conciliation Resources 2010, 8).
The peace agreement between Sudan and Uganda as well as the
amnesty put significant pressure on the remaining LRA fighters who
responded with an increase in violence. Public statements now became
rare, and no more manifestos emerged until after the ICC indictments
and the commencement of the Juba peace talks in 2006. By December
2001, the US government had added the LRA to its “B” list of “other
terrorist organizations” and by March 2002, the Ugandan government
had passed its own Anti-Terrorism Act, effectively retracting the
amnesty offer (Dolan 2009, 53). Following the peace accord with
Sudan, the Uganda army received permission for a one-month military
campaign to destroy LRA camps in Southern Sudan (Human Rights
Watch 2006). Aided by US-sponsored training efforts, “Operation
Iron Fist” commenced in March 2002, but once again failed to fully
Transnational diffusion & the Lord’s Resistance Army
133
defeat the rebel group. LRA soldiers now pushed back into Northern
Uganda, where insecurity increased.
By 2003/2004, the number of internally displaced in Northern
Uganda had increased to about 1.7 million, from an estimated
400,000 in 2000. The number of humanitarian NGOs operating
around Gulu had risen from five in 1996 to more than 60 (Dolan
2009, 239). In the meantime, Kony and his rebels had established new
camps in Garamba National Park in the eastern part of the Democratic
Republic of Congo. After repeated failures to defeat the LRA militarily,
Museveni now called on the newly established ICC to investigate the
LRA leadership for its atrocities. Following the ICC indictments in
2005, a third shift in the framing of the conflict occurred, which is
discussed below following the sections on resource mobilization, and
targets and strategies.
Transnational mechanisms help explain shifts in the framing of LRA
violence that go unnoticed by alternative accounts that typically reduce
the rebels to an irrational “cult.” The initial shift in framing took place
as Kony and other aspiring rebel leaders combined a desire to challenge
the marginalization of Northern Uganda with a claim to be the true
representatives of an Acholi community morally weakened and misled
by traditional elders. The justification for attacking Acholi civilians
had roots in missionary teachings and was understood as a necessary
“cleansing” of a failed community. The active translation of a Christian ideology into the local context through processes of relational
diffusion allowed Kony to construct a rationale for violence, recruit
fighters, and turn them against their own ethnic group.
In the mid to late 1990s, a second shift of framing took place when
the LRA was increasingly challenged on its human rights record and
use of fundamentalist Christian terminology. The framing now became
more directly driven by the influence of external actors without a direct
presence in the region, pushing the LRA into defensive statements and
a denial of responsibility for atrocities. This non-relational diffusion of
ideas produced a response of written statements and manifestos projecting a secular character and a mainstream political agenda with
frequent references to human rights and democracy.
The main challenge to including such transnational mechanisms into
the analysis of LRA violence is captured by the idea of “rebels without
a cause” and the claim of a new type of non-ideological, “chaotic
warfare” spreading across Africa after the end of the Cold War
134
Hans Peter Schmitz
(Reno 2009, 9). But these claims cannot account for the shifts in
framing identified here because they reject the possibility of any significant changes in the behavior and motives of Kony. In this view, the
manifestos are not authentic or represent mere “window-dressing” and
changes in framing are solely driven by military imperatives and material incentives on the ground.
While local grievances and conditions clearly shape much of the
conflict, transnational mechanisms take a central place during crucial
shifts in framing, strategies, and resource mobilization. Understanding
their impact is imperative for scholars and policy-makers interested in
shaping a strategy designed to exploit the political opportunities
offered by a rapidly changing situation. This insight is confirmed by
the deliberate efforts of the Ugandan government not to end the
violence and protect civilians and instead to cast the rebels as “backward” and isolated. And while the manifestos were erased from the
official discourse about the conflict, Finnström’s work shows that they
were common knowledge among the local population and shaped their
views of the rebellion. One may even argue that despite the glaring
gaps between the rhetoric of the manifestos and the actions of the LRA,
the manifestos accomplished the goal of reminding the local population of the failures of the government (Finnström 2008, 129). Finally, if
changes in the framing of the violence during the 1980s and 1990s had
been taken more seriously by many international observers, it may
have helped in anticipating better how the LRA would respond to the
2005 ICC indictments. The subsequent Juba peace process exposed
significant internal divisions within the LRA, marking a third shift in
the framing of the violence (discussed in the final empirical section
below).
Repertoires of violence: changing levels of atrocities
Taking seriously the shifts in the framing of the violence by the LRA
exposes the persistent disconnect between frequent demands for
human rights and democracy and the atrocities committed. The main
empirical puzzle for this section is to assess the importance of transnational mechanisms shaping the targeting and strategies of the LRA,
in particular its use of abductions and atrocities committed against
local residents of attacked villages. Apart from regularly engaging
military forces, the LRA used the attacks on villages to steal food
Transnational diffusion & the Lord’s Resistance Army
135
supplies and forcibly recruit new fighters. While activists and international media have almost exclusively focused on the LRA’s abduction of children, more recent studies based on interviews and surveys
(Allen and Schomerus 2006) have shown that such a view often
“infantilizes the movement, suggesting a lack of accountability and
making it appear a less than viable partner in peace negotiations. This
has been a problem during the Juba peace talks” (Schomerus 2007,
16). Child abductions played a relatively minor role in the overall
strategy and success of the LRA and the main recruits were young
adults capable of receiving effective military training (Blattmann and
Annan 2010).6 Children were frequently “released after having done
duty as porters” (Schomerus 2007, 16). The LRA was successful
because it had a strong, centralized command, a virtually unlimited
supply of weaponry (see section on resource mobilization below), and
enough recruits to sustain a military campaign across four nations for
almost 25 years. While child abductions were important for the LRA,
focusing exclusively on them inhibits an analysis of the significant
shifts in the strategies of the LRA, in particular, the rise in atrocities
early in the military campaign and the apparent abandonment of some
of the worst atrocities in the late 1990s.
Ignoring the political and social agenda of the rebels allowed activists to avoid the complexity of the situation and mobilize moral
outrage. As a result, even well-informed observers concluded that the
LRA was “not motivated by any identifiable political agenda, and its
military strategy and tactics reflect this” (International Crisis Group
2004, 5). In this view, the rebel group not only lacked rational goals,
but its selection of targets and strategies was largely random. In contrast, this section will elaborate why and how we can observe particular
shifts in the targeting and strategies used by the LRA and which transnational mechanisms redirected strategizing and enabled different levels
of organized violence. Two questions are central to the inquiry. First,
why did the LRA turn against the population in Northern Uganda
which it claimed to defend? How did it adopt practices of torture
and maiming in its brutal campaign against civilians in Northern
Uganda and three other nations? Second, while the LRA used
6
Scholars with field experience interviewing former rebel fighters point out that
NGOs and international media trumped up the issue of child abductions, when
such actions were actually “a lesser part of a far larger pattern of adult abduction
and a degree of voluntary recruitment” (Dolan 2009, 90).
136
Hans Peter Schmitz
abductions consistently throughout its entire military campaign, why
were mutilations of civilians most prevalent between 1991 and 1996
(Dolan 2009, 79), and later on played a much less significant role?
Lack of governmental control over borders and the ability of LRA
fighters to move easily across them enabled a strategy of increased
brutality against the local population. But the crucial transnational
mechanism was not simply crossing a border, but what the LRA
fighters learned and emulated while being exposed to an ongoing civil
war while establishing camps in southern Sudan in the early 1990s.
Similar to Bakke’s emphasis on transnational insurgents shaping the
strategies of Chechen rebels (Bakke, this volume), LRA fighters learned
some of their tactics after they were forced to flee northern Uganda and
were thrown into a decades-long, brutal civil war in southern Sudan.
One Sudanese politician said that she felt the LRA changed its behavior after
moving to Sudan, partly because they were encouraged by Khartoum to
commit atrocities, but also because the already brutal environment increased
the spiral of violence. She said, “They [the LRA] kill people and hang them
up, so when local people come, they don’t touch anything. I think the LRA
learned these things from southern Sudan” (Schomerus 2007, 20).7
But it would be short-sighted to look only as far back as the late 1980s in
accounting for the origins of these atrocities. The violence experienced
and reproduced by the LRA can be traced back to British colonial policies
which had given preferential treatment to ethnic groups in the South and
reduced Northern Uganda to a recruitment pool for foot soldiers. At
independence in 1962, Milton Obote, a Langi from the North became
the first President, immediately relying on the military strength created by
the British among his own ethnic group. One of his generals, Idi Amin,
would dispose of Obote in 1971 and expand military rule, leading the
nation into more than a decade of bloodshed and civil war ending only in
1986. While Obote came back to power in 1980 and temporarily
reinstated the North’s role in national politics, Museveni’s military victory in 1986 signaled a profound crisis of Acholi identity and opened up
opportunities for new leadership developing violent strategies.
Alice Auma and Joseph Kony relied on a mix of spirituality and
references to religion to offer the Acholi a form of redemption and
7
The quote is from an unnamed government official based in Juba/Sudan, who
was interviewed by Schomerus in November 2006.
Transnational diffusion & the Lord’s Resistance Army
137
return to their identity following their “near destruction through using
modern methods and structures” (Jackson 2002, 49). Both developed
complex rituals and introduced a world of “spirits” to heal and cleanse
individuals, including preparing fighters for battle. And each insisted
that war was a form of healing and that “Acholi society had to be
purified by violence” (Allen 2006, 40). While Auma’s Holy Spirit
Movement did not attack civilians during its brief military campaign
from 1986 until its defeat in 1988, the LRA turned against its own
population as a result of the particular framing of the crisis of Acholi
society and its violent experience in southern Sudan. Kony is cited
defending such practices during the 1994 peace negotiations:
If you pick up an arrow against us and we ended up cutting off the hand you
used, who is to blame? You report us with your mouth, and we cut off your
lips. Who is to blame? It is you! The bible says that if your hand, eye or
mouth is at fault, it should be cut off. (Allen 2006, 42)
As the LRA was engaged in military combat by Ugandan forces, the
practice of brutal attacks on civilians was further reinforced. Starting
around 1991, mutilations became a way of deterring local collaboration with the government troops and many recruits were forced to
commit atrocities in order to make it “difficult for them to return
home” (Allen 2006, 42). Once the LRA received substantial material
support from the Sudanese government and was also militarily strong
enough to regularly raid villages and trading centers for supplies,
support by the local population became less crucial and abductions
served as a main recruitment strategy. The LRA abducted younger
adolescents “because not only were they easier to indoctrinate, but
Kony had given up on the adults. The youngsters were to form the core
of a new Acholi identity” (Jackson 2002, 49).
Why did the LRA then begin to limit the practice of regularly
mutilating civilians after 1996? Throughout its entire military campaign, the LRA consistently relied on raids and abductions to meet its
material needs and resupply its ranks of fighters. The rebels also
continued to regularly kill civilians indiscriminately during their raids.
But some of the particularly gruesome attacks on civilians began to
diminish in the mid 1990s and reports about the activities of the LRA
contained fewer references to maiming and torturing, while continuing
to highlight abductions and looting. Concomitant with the changes in
framing discussed above, the increased exposure to human rights
138
Hans Peter Schmitz
mobilization as well as the stepped-up activities of the diaspora community in the mid 1990s represent a form of weak transnational
“shaming” transported largely through indirect mechanisms of nonrelational diffusion. As the LRA found decreasing support from the
Sudanese government in its proxy war against Uganda, the rebels’
strategic freedom diminished. And after the ICC indictments signaled
a success of the human rights mobilization, local actors such as the
South Sudanese government, but also international relief organizations
were able to offer material rewards to the LRA in exchange for not
targeting civilians.
Transnational mechanisms offer unique insights into the LRA’s
strategic behavior, in particular, why it turned against civilians, committed atrocities, but then also dropped some of these tactics again
from its repertoire. Processes of learning and emulation in the context
of the southern Sudanese conflict played an important role in the initial
shift of increased brutality against the local population in northern
Uganda. While Kony had developed a rationale for why targeting his
own ethnic group was necessary, the atrocities intensified only after the
LRA moved to Southern Sudan. The geography of the border regions
offered options for retreat and encouraged abductions, thus clearly
establishing an observable effect of transnational mechanisms largely
neglected in the literature. Throughout the conflict, the LRA behaved
opportunistically in targeting civilians and frequently moved across
borders, initially from northern Uganda to Sudan in the mid 1990s,
back to Uganda in 2002, to the DRC in 2005, and again to southern
Sudan and, most recently, to the CAR in 2008.
A second shift in the LRA’s strategic use of violence occurred in the
mid 1990s after the Acholi diaspora became more active in promoting
a peaceful solution and the first major human rights reports and news
about atrocities reached Western audiences. Abductions and raids
continued, but some of the methods of maiming civilians became less
prevalent. It would require additional research to determine if the
mechanism at work here was more direct relational diffusion through
personal networks between rebels and diaspora groups or if this
change in repertoires of violence was caused by the more indirect
influence of human rights mobilization focusing on the LRA. No
matter if either or both mechanisms were at work, the external engagement was in many ways counterproductive to the goal of ending all
violence against children and adults. Indeed, viewing the LRA as an
Transnational diffusion & the Lord’s Resistance Army
139
irrational “cult” undermined more promising efforts at relational diffusion, which would have taken seriously the underlying grievances of
the local population.
Resource mobilization: governmental interests and small
arms diffusion
In the evolution of the LRA and its violent struggle, different resource
streams and varying financial support shaped strategies and goals. But
access to weaponry and other resources never limited the LRA and its
military strength. From the early 1990s until about 2002, the rebel
group received resources and training from the Sudanese government,
but it could also rely on some support from members of the Acholi
diaspora community. Moreover, its regular cross-border raids and
abductions of children, youths, and adults made it quite self-sufficient.
Over time, the LRA relied on a wide array of external resources and
supporters, but looting villages and abducting civilians to carry the
supplies back to its camps represented a key activity ensuring survival
(Dolan 2009, Annex A: Testimony of LRA soldier). After 2006 and the
beginning of the Juba peace talks, the LRA also temporarily received
official humanitarian aid delivered by the Catholic charity Caritas and
others, while local governments paid the LRA not to attack their
territory (Human Rights Watch 2006). As recently as May 2009, the
Ugandan Army reported intercepting trucks driven by Belgian nationals in southern Sudan carrying high-tech supplies for the LRA into the
DRC (Egadu 2009).
Research on diasporas has long explored the material and ideational
support these communities can provide in supporting violent rebellions
(Adamson 2004; Lyons 2006). But this kind of support was likely less
relevant in the case of the LRA. Diaspora resources may have been part
of the arms supply and other material support, but most observers
agree that the links between the leadership fighting on the ground and
the often self-appointed political representatives were tenuous at best.
It is difficult to obtain reliable information on the extent of resource
support for any given rebel group, especially looking back at almost
25 years of violent conflict. The most important resource – access to
weapons and ammunition for combat – was never an issue for the LRA
or any other rebel group operating in this area. A survey completed by
Nairobi-based Larjour Consultancy in 2001 found that an AK-47 sold
140
Hans Peter Schmitz
for about USh 25,000 (approximately USD 12) in the border regions
(cited in Van Acker 2004, 345). This means that relational diffusion
expressed in small arms supply by private dealers was sufficient to
allow the LRA to continue the violent struggle.
The first significant shift in resource mobilization took place when
the LRA relocated to southern Sudan in the early 1990s and began
receiving military equipment from the Sudanese government. The
weapons supply became more sophisticated and “the UPDF often fell
behind in the quality of their equipment” (Schomerus 2007, 41). Until
2005, the Sudanese government used the LRA and its military capabilities in its war against the SPLA, but also in the Darfur region. LRA
fighters would regularly fight alongside regular Sudanese forces against
the SPLA and their camps were based in the Juba area close to army
installations (Vinci 2007, 340). Kony and other commanders regularly
visited Khartoum and this resource support allowed the LRA to engage
in the kind of cross-border raids into Northern Uganda described
above (Schomerus 2007, 24; Dolan 2009, 82). While Kony’s particular
framing of the conflict offered a rationale for attacking civilians in
Acholi land and elsewhere, transnational geographical opportunities
and the transnational diffusion of resources enabled implementing
those violent strategies.
Stockpiles were easy to hide in the border regions and for its
communications the LRA used radio equipment and satellite phones
supplied by supporters. The supply of arms in the region was so
plentiful that the LRA did not have to engage in large-scale arms
trafficking, but frequently traded weapons to locals for valuable intelligence or money. Moreover, although the atrocities committed by the
LRA eventually ended the support by the Sudanese government
and also limited its ability to raise resources among the diaspora
(Schomerus 2007, 13), there is no indication that any of these issues
adversely affected its military strength.
After the Carter Center brokered a peace agreement between Sudan
and Uganda in 1999 explicitly calling for an end to support of rebel
groups, official supplies likely still reached the LRA, which continued
to operate from southern Sudan until 2002. While the Carter Center
had succeeded in getting a peace accord between two governments, the
negotiations failed to include any of the rebel groups such as the LRA
(Dolan 2009, 97). When the Sudanese government in 2002 allowed the
UPDF to pursue the LRA on its territory (“Operation Iron Fist”), the
Transnational diffusion & the Lord’s Resistance Army
141
violence was pushed back into northern Uganda, the number of internally displaced increased dramatically, and the LRA leadership ultimately found another safe haven in the Garamba National Park in the
DRC. As one security expert interviewed in 2006 put it: “Why is the
UPDF in Sudan? To make sure the LRA is not destroyed” (Schomerus
2007, 29). The other reason was (as was the case in the DRC in the late
1990s) that UPDF control over this territory allowed for Ugandan
soldiers to profit from the exploitation of natural resources such as
teak.
While some members of the Ugandan government and military
developed an interest in sustaining a low-intensity conflict with the
rebel group, the key resources for the LRA were increasingly no longer
related to the presence of those sustaining the violence (e.g. Sudanese
government support), but the absence of policies and capacities at the
international level to effectively protect civilians and address the underlying local grievances. The LRA atrocities became a public relations
resource for Museveni distracting from his dismal human rights record
and abuses of power. Knowing that the LRA was never a real threat,
his regime simply had to ensure that it “bolstered an appearance of
doing the right thing” (Dolan 2009) while perpetuating an image of
Kony as an insane cult leader.
The resource base of the LRA shifted after the Sudanese government
ended its support and again after the opening of the Juba peace
negotiations in 2006. What carries over from the earlier period is the
LRA’s emphasis on autarky and its continued use of raids and
abductions for survival. The supply of small arms also remains largely
in place, with no discernible decrease in the availability of such
weapons. With the commencement of peace negotiations, what does
change is the introduction of material rewards offered by local governments and channeled through international humanitarian groups.
Since 2005, government officials in southern Sudan regularly paid off
the LRA in exchange for not attacking their territory and population.
The LRA was also regularly “paid” to attend peace negotiations
(Johnson 2009) and even complained in 2007 about “not being
granted USD 2 million for consultations in northern Uganda.” These
rewards play a highly ambiguous role in actually accomplishing the
goal of ending violence because they simultaneously constitute “an
incentive to keep the process going and a disincentive to streamline it”
(Schomerus 2008, 96). Similarly, aid provided by NGOs also injects
142
Hans Peter Schmitz
new resources into conflict zones. Lack of coordination across NGOs
and humanitarian principles of neutrality offer rebel groups additional
“opportunities . . . to exploit these resources” (Reno 2009, 12).
The supply of small arms and support by the Sudanese government
played an important role in the context of lawless border regions. But
the LRA stands out as a rebel group sustained primarily by a selfsufficient strategy of mobilizing resources using looting and
abductions. Its military power was certainly significantly diminished
after 2002, but not necessarily its ability to create widespread insecurity and displacement of civilian populations. As the conflict progressed,
the importance of resources directly made available to the LRA diminished while the knowledge, capacities and resources absent at the
international level become increasingly relevant in sustaining the cycle
of violence. By not addressing underlying grievances and the complex
motives of individual fighters, the ability of the LRA to sustain its
violent campaign only diminished marginally. Only after the amnesty
offer went into force and the ICC indictments were announced did the
incentives for the rebels change – and the most recent shift in the
framing of the conflict occurred.
Frames of global justice and peace, 2003–2011:
ICC indictments and Juba negotiations
Museveni’s 2003 decision to involve the ICC and the 2005 warrants
issued for five LRA commanders fundamentally altered the framing of
the conflict. While transnational human rights activists defend the
indictments as a necessary step against impunity, many observers with
field experience in the region tend to be critical of the indictments and
argue that they undermine negotiations for a peaceful settlement of the
conflict (Allen 2005; Branch 2007). Since the establishment of the ICC,
scholars have debated vigorously the actual impact of the prosecutorial
approach on ongoing conflicts as well as post-conflict political developments (Kastner 2007; Vinjamuri 2010).
The ICC proceedings against the LRA were preceded by two decades
of transnational human rights mobilization, including the emergence of
child soldiering as a new major cause for global human rights activism.
In these campaigns, the violence in the DRC/Sudan/Uganda region
became a prime argument for the establishment of the ICC and activists insisted that such a court would deter future crimes of this
Transnational diffusion & the Lord’s Resistance Army
143
magnitude. For Yoweri Museveni, these global campaigns offered
another opportunity to advance his own political agenda, following
on the peace deal with Sudan in 1999, the never fully implemented
Amnesty Act of 2000 and “Operation Iron Fist” in 2002 which had
failed to capture Kony. In its referral application to the ICC, the Ugandan
government denied that the LRA was a legitimate belligerent under
humanitarian law (Government of Uganda 2003, 22); it was thus “befitting both from a practical and moral viewpoint to entrust the investigation and prosecution of these crimes to the Prosecutor of the ICC”
(Government of Uganda 2003, 14). The document reinforced the dominant discourse, claiming that “the LRA has not had a coherent ideology, a
rational political agenda, or any form of popular support” (Government
of Uganda 2003, 6). By referring the situation to the ICC, the Ugandan
government continued a deliberately incoherent approach of alternating
between offering amnesty and reconciliation with new military operations and other threats (Dolan 2009, 100).
When the ICC announced the indictments in 2005, international
reaction was overwhelmingly positive, while skepticism prevailed in
Uganda. Save the Children/Uganda issued a statement expressing concern that “the LRA leadership might apply even more strict discipline
to prevent witnesses from escaping” and predicted that the intervention would lead to “even more suspicion and distrust” (Allen 2006,
83–84). The indictments yet again raised the profile of the conflict in
the international media and provided the backdrop for the extensive
peace negotiations that took place from mid 2006 until 2008.
As a result of the 2002–2003 military “Operation Iron Fist” in
southern Sudan, the LRA was now on the move again, filtering back
into northern Uganda and leaving hundreds of thousands displaced
(Allen 2006, 72). As the humanitarian crisis worsened, international
attention and pressure for a negotiated settlement increased. In January
2004, the US Congress passed the Northern Uganda Crisis Response
Act, which for the first time implicitly criticized the Ugandan government and military for its own human rights violations and lack of
serious effort to negotiate an agreement. A national referendum held in
Uganda in July 2005 led to the reintroduction of multiparty politics
and Museveni’s political space diminished as he was preoccupied with
winning the elections in February 2006.
Pressure not only mounted on Museveni, but also the LRA and its
leadership (Schomerus 2008, 94). The military campaign had taken its
144
Hans Peter Schmitz
toll, visible in a significant jump in the number of LRA soldiers
accepting the amnesty offer in 2003 and 2004. Several reception
centers run by international NGOs such as World Vision reported
receiving about 5,000 former abductees and fighters by late 2004.
Around the same time, members of the diaspora succeeded in mobilizing donors and Betty Bigombe was brought back to Sudan to facilitate
talks once again. The 2005 Comprehensive Peace Agreement between
the Sudanese government and the SPLA led to the formation of the
semi-autonomous government of Southern Sudan and the LRA was
faced with three options: “withdraw from Sudan, declare war on the
SPLA, or engage in negotiations” (Schomerus 2007, 34).
After a series of meetings, the Vice President of the new South
Sudanese government, Riek Machar, brought both parties to the negotiation table in July 2006. Before the talks had begun, Machar was
filmed handing US$20,000 to Kony; he was quickly accused of bribing
the LRA to remain peaceful and show up. With the advent of the Juba
peace talks, the LRA/M established a more professional information
policy, expressing demands ahead of the negotiations and denying
responsibility for attacks (Schomerus 2007, 16). In its opening
remarks, the LRA/M delegation insisted once again on having a political agenda, stating: “Your Excellency, over the years, it has been
suggested . . . that the LRM/A has no political agenda . . . Failure of
the LRM/A to have access to the mass media to express its political
agenda in intellectual form does not mean the lack of it” (Finnström
2008, 127). A month before the first Juba talks, Kony had granted rare
interviews to international media where he claimed that LRA manifestos were widely available in Uganda and elsewhere. After being
asked why the LRA had failed to get its political demands disseminated
more widely, Kony replied: “All things [all information comes] from
Museveni’s side or from some other people, because I do not have
proper propaganda machineries” (Schomerus 2007, 14). In the same
interview, Kony acknowledged atrocities by stating that “I cannot say
that we are fighting clean war [or that] Museveni is fighting dirty war,
that one is difficult to say. Because a clean war is known by God only”
(Schomerus 2007, 15). But in another interview with a BBC reporter,
he flatly denied being responsible for any atrocities (Farmar 2006). At
a parallel meeting between LRA leaders and a large delegation of
Acholi civil society leaders from northern Uganda and southern Sudan,
Kony and others apologized for atrocities committed.
Transnational diffusion & the Lord’s Resistance Army
145
The initial agenda of the negotiations covered the establishment of a
ceasefire, political solutions, justice/accountability, and demobilization
and reintegration. While a Cessation of Hostilities agreement, signed
on August 26, guaranteed safe passage for LRA soldiers to designated
assembly areas (Government of the Republic of Uganda/Lord’s Resistance Army/Movement 2006, section 5), both sides violated the agreement and attacks between the UPDF and LRA remained common.
Comprehensive written agreements emerged on all issues, including
“political solutions to the situation in Northern Uganda” (May 2007),
“accountability and reconciliation” (June 2007) and the most difficult
issue of “disarmament and demobilization” (February 2008). During
the entire time period, LRA leaders repeatedly demanded an end to the
ICC investigations. In December 2006, Kony stated:
The international justice system is that if you are weak, the justice is on you.
For the time being, they think me, I am weak . . . Same with Taylor, when he
was in power nobody thought of justice. If you want to remain safe from the
ICC, you must fight and be strong (translated into English from Luo, cited in:
Schomerus 2008, 95).
The agreement on “accountability and reconciliation” signed by the
LRA and the government on June 29, 2007 sought to assure LRA
fighters that criminal and civil proceedings would take place in Uganda
before a special War Crimes Division to be established at the High
Court (Otim and Wierda 2010). Despite the significant progress in
negotiating the separate parts of the peace agreement, the negotiations
ultimately exposed deep divisions within the LRA’s military wing.
While Kony appointed all of the LRA/M representatives in Juba,
significant divisions existed not only between diaspora representatives
and the rebel group, but also within the leadership of the LRA. When
Kony’s second-in-command, Vincent Otti, apparently pushed in 2007
for a greater commitment on the part of the rebels, he was executed
and the negotiations further delayed (Otim and Wierda 2010, 5).
Despite the significant upheaval within the LRA ranks, the final
agreement was ready for signature in March 2008. In April and
September 2008, Kony twice failed to appear at a scheduled signing
ceremony for the Final Peace Agreement, citing both times the ICC
warrant as a reason. However, the ICC indictment only served as a
cover for deep divisions among those negotiating in the name of the
LRA. First, and most importantly, those speaking for the LRA “were
146
Hans Peter Schmitz
not the key decision-makers” and “things said in Juba would be
interpreted differently in the field or in Garamba Park” (Conciliation
Resources 2010, 18, interview with Julian Hottinger). Second, the
Ugandan government had been pressured into the talks and its commitment to a peaceful solution was limited. Museveni and his military
commanders were content with keeping the LRA physically out of
northern Uganda. Third, the negotiations exposed the lack of capacities at the United Nations level, where a department with no mediation
experience (UN Office for the Coordination of Humanitarian Affairs,
UNOCHA) was put in charge of the process (Conciliation Resources
2010, 19, interview with Jan Egeland).
The Juba process ended without Kony’s signature and by mid
December the militaries of Uganda, southern Sudan and the DRC –
under the direction of the US military command AFRICOM –
launched “Operation Lightning Thunder” (Johnson 2009; Schomerus
and Tumutegyereize 2009; International Crisis Group 2010). At the
same time, the situation in northern Uganda improved markedly, and
no major incidence of violence has been reported since the Cessation of
Hostilities agreement in 2006. About one million of those forced into
camps for the past two decades were able to return to their homes,
while another million remained in the camps for the internally
displaced or were transitioning into temporary facilities (Justice and
Peace Commission of the Archdiocese of Gulu 2008). In May 2009,
Kony moved his rebel group north from Garamba National Park to the
Central African Republic (CAR), extending violence into yet another
border region (International Crisis Group 2010).
The intervention of the ICC has significantly altered the conflict and
the incentives for the participants involved. Both the “shaming” pressure of transnational NGOs and the threat of prosecution have had an
impact on the LRA and especially the individual fighters. During 2003
and 2004, low-ranking members of the rebel group took advantage of
the amnesty offer, while the LRA experienced significant internal strife
as it participated in the peace negotiations. But this indirect effort to
target the LRA through principled mobilization and prosecution fundamentally failed to engage effectively with the local realities not only
of many fighters, but also the communities affected by the violence.
Effective socialization to end the violence would require a more relational approach that abandons the dominant frame of viewing the
LRA as a cult. Instead, efforts of the international community were
Transnational diffusion & the Lord’s Resistance Army
147
uncoordinated and often contradictory, including bribery, threats of
prosecution, and a prevailing rhetoric of military power.
Conclusions
Understanding the significant shifts in the goals, strategies, and resource
mobilization of the LRA requires an in-depth analysis of how violence is
shaped by external influences and the transnational mechanisms facilitating interactions between the local and the global. The framing and
justification of LRA violence shifted as a result of many, and often
hegemonic external efforts to foster learning and emulation. The evidence suggests that social adaptation in the context of the Southern
Sudanese conflict provides a compelling explanation for why heinous
attacks on Acholi civilians increased during the early 1990s. Subsequent
efforts by international human rights organizations to “shame” the LRA
for its atrocities as well as an increased effort by the diaspora community
to end the violence preceded an almost complete stop in the deliberate
use of maiming and other extremely violent practices. But these socialization efforts based on non-relational diffusion mechanisms had little to
no effect on continued abductions and the wanton pillaging of villages.
The primary reason for this failure to end the violence was the reliance
of human rights activists on abstract norms at the expense of a more
meaningful engagement with local grievances.
In the context of resources, transnational mechanisms came in many
different forms. Small arms supply by private sources was important
throughout the conflict, except during the time the Sudanese government directly supported the LRA during the 1990s. Its strategy of selfsufficiency made the LRA independent from external pressures,
although diaspora and other resources likely expanded its military
capacity. The major shifts in the resource supply did affect the rebel
movement, but often those shifts were countered by the availability of
alternative sources, preserving the basic disruptive capacity of the
LRA. For example, when the Sudanese military support ended and
the LRA began to access very different resources in return for sitting at
the negotiating table, it had no major impact on its military strategy
and attacks on civilians, which resumed as soon as the negotiations
were over.
Against the prevailing views of the LRA as an incoherent and
irrational “cult,” this chapter showed how a largely isolated rebel
148
Hans Peter Schmitz
group interacts in sophisticated ways with its transnational environment through specific social mechanisms of relational diffusion,
emulation, and situational learning. But the conflict was most effectively perpetuated not by the presence of any of the mechanisms discussed here, but by the absence of an effective strategy based on
relational diffusion, which would have addressed the root causes of
this conflict. The dominant frame of viewing the LRA as irrational not
only privileged non-relational approaches and prevented any sustained
efforts to engage directly with the group, but also played well into the
Ugandan government’s political interests of sustaining a low-key violent conflict in the region. Without greater coordination of different
external efforts and by not taking seriously local grievances and the
transnational mechanisms at work, the international community is
perpetuating its repeated failures and neglecting a more promising
effort that would shift attention to the effective protection of civilians
and the resolution of the conflict.
|
6
Transnational advocacy networks,
rebel groups, and demobilization
of child soldiers in Sudan*
stephan hamberg
Introduction
In 2001, the Sudan People’s Liberation Army/Movement (SPLA/M),
while still fighting a war against the Khartoum government, demobilized what they claimed was their entire population of child soldiers
(International Crisis Group 2002 and UNICEF 2001a; see also Singer
2004). Why? Did transnational activists shame the SPLA/M into
demobilizing their child soldiers? Did the SPLA/M demobilize child
soldiers because they were losing the war, and such demobilization
would improve their moral position in future peace negotiations?
Alternatively, were they winning the war, and thus could afford to
reduce the number of soldiers in their ranks? Did the international
community, with the US in the driver’s seat, threaten or promise the
SPLA/M something in return for demobilizing the child soldiers? These
are all potential rival explanations and mechanisms for why a nonstate armed group would demobilize soldiers while fighting a war.
This chapter takes a different starting point to many of the empirical
studies in this book. Rather than focusing on the influence of diasporas
on rebel groups (Adamson, this volume), or how transnational insurgents influence rebel groups (Bakke, this volume), my chapter focuses
on whether transnational human rights networks impact rebel groups.
As Checkel (this volume) points out, the literature on transnational
advocacy networks (TANs) has focused on peaceful transnational
dynamics, but has rarely examined whether such networks are successful in reducing human rights violations during violent civil conflict.
This literature has also focused almost exclusively on how TANs can
influence state behavior, thus neglecting how transnational actors can
* Thanks to Jeffrey Checkel, Kristian Berg Harpviken, Kristin Bakke, Andrew
Bennett, Matthew Evangelista, Scott Gates, Elisabeth Wood, Martin Austvoll
Nome and the other project participants for their many helpful comments on the
chapter.
149
150
Stephan Hamberg
influence other non-state actors (both domestic and transnational). It is
these gaps my chapter addresses. More specifically, it seeks to examine
if and how transnational human rights advocacy networks can influence the behavior of rebel groups fighting in civil wars.
Rebel groups violate human rights on a daily basis by targeting
civilians, using rape as a weapon and recruiting child soldiers. Some of
the most serious human rights violations today are committed by nonstate armed groups (Policzer 2006). As most rebel groups do not have
formal legal and political status, the international community lacks
many traditional tools when trying to reduce human rights violations
by them. However, transnational human rights advocacy networks
have during the last decade increased efforts to change rebel group
behavior. Naming and shaming is the preferred tactic TANs use when
trying to influence state and rebel group behavior. However, few efforts
have been made to assess the efficacy of this tactic when it comes to the
behavior of rebel groups. In this chapter, I use the SPLA/M’s decision to
demobilize thousands of child soldiers in 2001 to test, among other
alternative explanations, the power of TANs and naming and shaming
as a mechanism to induce rebel groups to change behavior.
Even though child soldiers have been used in wars throughout
history, only recently has this phenomenon received attention by
policy-makers, non-governmental organizations (NGOs), international
organizations (IOs), media, and academics. NGOs and networks
of NGOs and IOs – so-called transnational advocacy networks – are
primarily responsible for the recent attention being paid to the issue of
child soldiers. Thus, TANs have been successful in bringing attention to
a problem they care about. They have also been successful in pushing for
the creation of international rules and conventions against the use of
child soldiers.1 These advocacy networks have been so successful that
we now speak of a norm against the use of child soldiers. However,
despite the attention the subject has received, the new international rules
and conventions against the use of child soldiers, and the more informal
norm against their use, the number fighting in ongoing conflicts remains
high. Indeed, the main explanation for why the total number of child
soldiers we observe is lower today than some years ago is that many
wars involving significant numbers of them have recently ended (Gates
1
See Achverina and Reich (2009) for a list of all international rules and
conventions against the use of child soldiers.
TANs, rebel groups, & demobilization of child soldiers
151
and Reich 2009). This raises an interesting puzzle: why, despite the
considerable effort by transnational advocacy networks, do we not
observe that rebel groups involved in civil wars demobilize child soldiers? And more specifically, why did the SPLA/M demobilize child
soldiers in 2001?
The answer is that transnational advocacy by US and Southern Sudanese civil society actors towards US politicians and the SPLA/M led the
rebel group to demobilize child soldiers to secure financial and political
support. What emerges from the empirics is a modified boomerang
pattern (Keck and Sikkink 1998), in which Southern Sudanese church
organizations linked up with US religious organizations, which in turn
put pressure on key US policy-makers to not only get involved in the
Sudanese civil war, but to explicitly support the SPLA/M’s cause. Rather
than a traditional boomerang pattern, where domestic groups seek international support to name and shame human rights violators, this case
shows that the New Sudan Council of Churches (NSCC) pressured the
SPLA/M from below, while simultaneously securing support for the rebel
group from the US government.
The remainder of the chapter is divided into three sections. First,
I discuss four different approaches to explaining the SPLA/M’s decision to demobilize child soldiers; I elaborate the causal mechanisms at
work and discuss specific testable implications for each. Second, I test
which explanation best explains SPLA/M’s decision to demobilize their
child soldiers. I conclude in the final section.
Rebel groups and child soldiers
Why does a rebel group choose to demobilize child soldiers? Any
attempt to understand whether transnational factors played a role,
must first examine whether domestic or internal reasons explain the
outcome of interest.
Most of the literature on child soldiers refers to demobilization as
any instance of a former soldier leaving the army, either voluntary or
not. However, if a soldier, whether a child or not, leaves an army
without the consent of his or her officer, he or she has not been
demobilized; rather, they have deserted.2 Demobilization, as used in
2
Desertion is normally associated with untrustworthy soldiers not committed to
fighting. Though this is often looked upon with negative connotation, I imply no
152
Stephan Hamberg
this chapter, refers to a decision by the top leadership of a military
organization to discharge soldiers from the unit and let them rejoin
society as civilians. In other words, demobilization only occurs if the
leadership decides that some troops can leave the armed group, and
not when soldiers leave without its consent. By defining a demobilized
soldier as any soldier who has left an armed group, we are likely to
overestimate the effect naming and shaming has on rebel groups, as we
would count both soldiers who have been asked to leave by the
leadership, as well as soldiers who have escaped an armed group they
have been forcefully recruited to join.
Conceptualizing demobilization as a decision by the leadership in
armed groups to discharge soldiers has several benefits. First and as
suggested above, it will increase the accuracy of the numbers of soldiers
who are actually demobilized by armed groups. Second, if one wants to
understand whether international pressure impacts rebel group behavior, one needs to look specifically at the leadership’s behavior and not
an individual soldier’s decision to desert or not.
Though we lack precise data on the child soldiers that serve, or have
served, in armed non-state groups, we know the number runs to
several thousands. It is also clear that the number of children serving
as soldiers increased significantly in the latter half of the twentieth
century (Singer 2006). As transnational advocacy groups began to
campaign against the use of child soldiers, scholars responded by
trying to explain the causes of child soldiering. Any attempt to explain
their demobilization must consider the causes of child soldier recruitment. For example, if the cause of the increased number of child
soldiers is poverty, then a reduction in the number of child soldiers
might be caused by a reduction in poverty. Thus, it is necessary to
examine the mirror image of the causes of child soldier recruitment.
The latter literature can be divided into explanations focused on the
supply of children and those examining the demand for them. To date,
most research has focused on supply-side arguments.3
Several supply-side explanations can explain the increase in child
soldiers during the last 20–30 years. Summarizing qualitative research
3
such thing here. Many soldiers, and especially child soldiers, have been forcefully
recruited and desertion in this case is more accurately described as legitimate
escape.
To my knowledge, Andvig and Gates (2009) are the first to separate the cause of
child soldiering into demand and supply based explanations.
TANs, rebel groups, & demobilization of child soldiers
153
on child soldiers, Ames (2009) distinguishes three categories of causes.
First are grievance factors: poverty, loss of parents, lack of economic
opportunity, abuse at home, ethnicity, and political beliefs. Second are
inducement factors: improved income opportunities, glory, and promises of future payoffs. Third are solidarity factors such as group cohesion, village networks, and friends (Ames 2009, 15). Peter Singer
(2006, 37–56) argues that the increase in child soldiers can be
explained by high poverty levels, large numbers of orphaned boys,
new weapons technology, and increased availability of small arms,
especially after the end of the Cold War and the break-up of the Soviet
Union. Achverina and Reich (2009) propose an additional supply-side
explanation. They argue that Singer’s explanations fail to explain why
some states and rebel groups have fewer child soldiers than others, and
argue that variation in child soldier recruitment is better explained by
whether armed groups have access to weakly secured refugee camps.
According to Achverina and Reich, it is “the degree of vulnerability of
children in refugee/IDP camps that ultimately explains their participation rates” (2009, 57).
This list of factors all focus on the supply-side. Children with more
economic opportunities do not become child soldiers, secure refugee
camps reduce the opportunity armed groups have in terms of recruitment, and fewer small arms make children less useful as soldiers, etc. It
is clear that all of these factors can, and probably do, explain the
increased number of child soldiers the world experienced in the last
three decades. However, and this is the key point, none of these factors
can explain variation in demobilization of child soldiers. For example,
better protected refugee camps can explain why we do not see new
recruitment of child soldiers, but it cannot explain why armed groups
choose to demobilize child soldiers. To examine whether the causes of
child soldier recruitment mirror the causes of child soldier demobilization, one must look at demand-side explanations.
Compared to the literature on supply-side explanations, relatively
little has been written on the demand-side. Demand-side explanations
assume that child soldiers and adult soldiers are substitutable goods
(Andvig and Gates 2009). With technological innovations in small
arms and a large amount of AK-47 machine guns available on the
market, children are relatively efficient soldiers, and thus armed groups
might prefer child soldiers if they are cheaper to recruit and maintain
than adult soldiers. Andvig and Gates (2009) write that child soldiers
154
Stephan Hamberg
can be beneficial to armed groups in several ways. Children are easier
than adults to recruit – both forcefully or voluntary – and indoctrinate
(see also the socialization mechanism discussed in Chapter 1). They
require less monetary benefits to fight, are more likely to obey orders,
desert less often, and are often less fearful than adult soldiers. As rebel
groups are relatively poor, we should expect these groups to use the
cheapest form of soldier. Andvig and Gates use the example of
Mozambique and the Renamo rebel movement to show this dynamic.
As long as Renamo was supported by South Africa, the group could
pay adult soldiers to fight for them, but as South Africa withdrew its
support, Renamo significantly increased its use of forced recruitment
and of child soldiers (Andvig and Gates 2009, 85). Whereas supplybased explanations cannot be used to explain demobilization of child
soldiers, these demand-based explanations and market-based mechanisms can.
As Andvig and Gates (2009) show in the case of Mozambique,
support from outside actors can decrease a group’s demand for child
soldiers. In the case of demobilization, one can argue that a group
might demobilize child soldiers if it receives financial support from
outside actors. An armed group may also acquire new and larger
revenue streams through access to natural resources or aid. Following
the logic of this argument, one would expect that rebel groups who
increase their revenues would demobilize their child soldiers.
Demand for child soldiers can also vary depending on dynamics on
the battlefield. If an armed group is winning the war, its demand
for soldiers will decline and we should expect them to demobilize
soldiers – child or not – regardless of age. This is particularly true for
insurgencies that need the support of the population after the war;
demobilizing child soldiers might improve their image both at home
and abroad. If battlefield dynamics and the need to improve their image
explain SPLA/M’s decision to demobilize child soldiers, we should
expect to see them go to great lengths to publicize their actions.
How might transnational human rights mobilization affect these
demand-side dynamics focused on material support and image? Here,
a natural starting point is Keck and Sikkink’s (1998) boomerang
model, in which they argue that domestic rights organizations link up
with international human rights groups, which – in turn – put pressure
on the violator from the outside. This pressure takes the form of
persuasion and socialization, that is, “reasoning with opponents, but
TANs, rebel groups, & demobilization of child soldiers
155
also bringing pressure, arm-twisting, encouraging sanctions, and
shaming” (Keck and Sikkink 1998, 17). Persuasion through reasoning
can certainly take place, but the key to the boomerang model is to
increase the material cost to the human rights violator, either through
material sanctions or moral leverage. On the latter, shaming plays an
important role; it insures the behavior of the violator is publicized and
thus “held up to the light of international scrutiny” (Keck and Sikkink
1998, 23). The logic behind shaming is that governments, even those
who are serial violators of human rights, “value the good opinion of
others” (Keck and Sikkink 1998, 23). Being viewed as a pariah state is
costly, not only in moral currency, but also materially. The goal of
shaming campaigns is to increase both the material and moral cost of
human rights violations.
Several studies have shown that the boomerang model in general,
and naming and shaming more specifically, successfully explain positive changes in human rights behavior (Keck and Sikkink 1998; Risse,
Ropp and Sikkink 1999a; Klotz 1995).4 Yet, as Keck and Sikkink
point out, certain conditions increase the likelihood of success. In
particular, success is more likely when shaming is about issues that
involve “bodily harm to vulnerable individuals, especially when there
is a short and clear causal chain” (Keck and Sikkink 1998, 27).
The use of child soldiers is a human rights issue that has received
significant attention during the last two decades; it also seems exceptionally well suited for a naming and shaming campaign as it is
specifically about bodily harm to vulnerable individuals. Thus, testing
whether the boomerang model in general and naming and shaming in
particular can explain the SPLA/M’s decision to demobilize child soldiers is imperative.
If the boomerang model and naming and shaming explains SPLA/
M’s decision, we should expect to see several things. These include
domestic civil society groups linking up with international or transnational advocacy groups; consistent shaming over time of the SPLA/
M; shaming campaigns directed at the SPLA/M’s use of child soldiers
specifically; and transnational advocacy groups calling for material
sanctions on the SPLA/M.
4
For a dissenting view, see Hafner Burton 2008; 2009; Hafner Burton and Ron
2009.
156
Stephan Hamberg
One potential explanation that has received relatively little attention
in the literature on human rights advocacy is the use of promises or
concessions (Hafner-Burton 2009; Katzenstein and Snyder 2009).
Whereas the boomerang model and the naming and shaming mechanism focus on increasing the cost of certain behaviors, using promises or
concessions would reward changes in behavior. Even though these
might seem to be similar or just different sides of the same coin, they
are in fact quite different. Most rebel groups fighting in civil wars have
little to lose. After all, they are willing to risk their lives for some more
or less specific purpose. Thus, being threatened with sanctions or
moral shaming has limited effects on their cost-benefit calculations.
Promises or concessions on the other hand, as long as they increase the
chances for an insurgency to achieve its goals, can significantly
improve the group’s expected utility without imposing many costs. If
the SPLA/M demobilized child soldiers because they were promised
something, we should observe that the child soldiers are demobilized
after the promise has been put forward, but before any actual
concessions.
In sum, to assess which of the explanations best explains the SPLA/
M’s decision requires that we carefully examine the testable implications of each mechanism. Table 6.1 summarizes the alternative explanations and the observable implications associated with each one.
The Sudanese Liberation Army’s decision to demobilize
child soldiers in 2001
In October 2000, in a letter to Carol Bellamy, the Executive Director of
UNICEF, the SPLA/M committed to demobilize all child soldiers from
its ranks; by August 2001, the SPLA/M had done so. This is the only
time a rebel group has demobilized most of its child soldiers while still
fighting a civil war. Considering the amount of effort human rights
organizations have put into reducing the use of child soldiers, this is an
excellent case to assess whether their work has paid off.
To assess which of the alternative explanations best explains the
SPLA/M’s decision, I collected both primary and secondary data, including
personal communication with policy-makers, scholars, and civil society activists. What emerged was a narrative that provides important
confirming evidence for the concessions mechanism. Interestingly, the
empirics also suggest that a modified boomerang model best explains
Table 6.1 Alternative explanations and testable implications
Observable
Explanation/mechanism implication I
Observable
implication II
Resource availability
mechanism
Did the SPLA/M get access Did the SPLA/M
to new resources
substitute child soldiers
with adult soldiers?
PRIOR to
demobilization of child
soldiers?
Battlefield dynamics/
winning war improve
image
Was the SPLA/M winning Did the SPLA/M publicize
the war?
their decision to
demobilize
internationally?
Boomerang/shaming
Did domestic human
rights groups link up
with international
groups?
Concessions/promises
Did the SPLA/M
Did someone promise
concessions to the
demobilize child
soldiers prior to
SPLA/M in return for its
receiving any
demobilization of child
soldiers?
concessions?
Did human rights
organizations
consistently shame the
SPLA/M?
Observable
implication III
Observable implication IV
Were shaming
Did transnational
campaigns directed at
advocacy groups call
for material sanctions
the SPLA/M’s use of
child soldiers
targeting the SPLA/M?
specifically?
158
Stephan Hamberg
how the SPLA/M was able to secure promises of financial and moral
support from the US government in return for a better human rights
record. Indeed, it was a process in which transnational advocacy,
information sharing, and persuasion played an important role. However, where the boomerang model predicts that naming and shaming
will be the key mechanism behind behavioral change, my empirics
indicate that they in fact played little role in the SPLA/M’s decision to
demobilize child soldiers.5
In what follows, I first give a brief overview of the civil war in Sudan,
and the origin of the SPLA/M and its use of child soldiers. I then show
how transnational advocacy, information sharing, and moral persuasion convinced US religious organizations and policy-makers to back
the SPLA/M in their war against the Sudanese government. I also show
that US policy-makers indirectly provided promises of material and
moral support to the SPLA/M in return for a better human rights
record.
The SPLA/M emerged in 1983, shortly after the beginning of the
second civil war in Sudan. The movement published its manifesto in
1983, but to understand its demands and goals, we need to examine
the underlying causes of the conflict. Both civil wars in Sudan have
generally been classified as a war between Arabs and Africans,
Muslims and Christians, or a war between North and South Sudan,
but as Johnson (2003) shows, these labels are simplistic. He argues that
the war(s) have been caused by the interaction of several factors. First,
there are the patterns of governance established in Sudan long before
independence. These patterns established “an exploitative relationship
between the centralizing power of the state and its hinterlands or
peripheries, mainly through the institutions of slavery and slave
raiding, creating groups of peoples with lasting ambiguous status in
relation to the state” (Johnson 2003, xviii). It is clear that these exploitative relations have taken place between Khartoum in the North and
South Sudan, but the patterns have also been replicated in the North
between Khartoum and Eastern and Western Sudan. Thus, it is not
only a North vs. South question.
Second, there has been the recurring introduction of particularly
militant brands of Islam and attempts to Islamize all of Sudan (Johnson
5
Schmitz, this volume, similarly finds that naming and shaming had a limited effect
on the behavior of the Lord’s Resistance Army.
TANs, rebel groups, & demobilization of child soldiers
159
2003, xviii). These attempts to introduce Islam and sharia to the
entirety of Sudan have severely limited the rights of non-Muslim
citizens. Third, the first two factors have led to great inequalities in
economic, educational, and political development between the center,
Khartoum, and other regions, especially Southern, Western, and Eastern Sudan. Finally, these combined factors have led to a “failure to
obtain a national consensus in either the North or the South concerning national unity, regional development, and the balance of power
between the central and regional governments” (Johnson 2003, xix).
The interaction of these factors led to the first civil war that lasted
from the early 1960s to 1972, and the second civil war that began in
1983, and hopefully ended with the signing of the Comprehensive
Peace Agreement between the government in Khartoum and the
SPLA/M in 2005.
The use of children or very young men as soldiers in Sudan has a
long history. As early as the 1870s, the Turco-Egyptian army enslaved
young boys, who at the ages of seven to ten, would work as “gunboys” for the army. These “gun-boys” started out by carrying the guns
for the soldiers, but, as they grew, they became soldiers themselves
(Johnson 1992, referenced in Human Rights Watch [HRW] 1994).
According to Johnson (1989), this practice has continued until recent
times and includes the SPLA/M.
Reports about the SPLA/M’s recruitment of child soldiers began to
appear in the early 1990s (Amnesty International 1991; HRW 1994).
According to these sources, the SPLA/M encouraged children to leave
for refugee camps in Ethiopia, where they would receive education. On
arrival, boys would quickly be separated from the other refugees and
housed in separate buildings, where they received both regular education and military training. In 1994, HRW reported that:
there were some 17,000 boys in these camps, where, according to SPLA/M
officers and the children themselves, they were given military training as
well as education. They were removed from the camps for military service
when the needs of the SPLA/M demanded.6 (HRW 1994, 7)
As with most literature on child soldiers, no exact number exists on
how many young people have fought for the SPLA/M. However, there
6
On the militarization of refugees in civil war, see also Harpviken and Lischer, this
volume.
160
Stephan Hamberg
is no doubt that the number is in the thousands. Indeed, the Coalition
to Stop the Use of Child Soldiers reported in 2008 that the SPLA/M
had demobilized nearly 20,000 child soldiers in the 2001–2006 period
(Coalition 2008, 319).
In October 2000, UNICEF Executive Director Carol Bellamy met
with the SPLA/M leadership in Sudan. During this meeting, the deputy
chairman of the SPLA/M, Salva Kiir Mayardit, guaranteed in writing
that the organization would no longer recruit soldiers under the age of
18, and that they would demobilize all soldiers under the age of 18.
According to UNICEF, Salva Kiir told Bellamy that “for the good of
our country we want all the children of southern Sudan out of the
military” (UNICEF 2000). In the period February 23–27, 2001, the
SPLA/M demobilized a total of 2,600 child soldiers ranging from 8–18
years of age (UNICEF 2001a). By August 2001, nearly 3,500 former
child soldiers had been repatriated to their home communities with the
help of UNICEF (UNICEF 2001b). In the following five years, another
17,000 child soldiers were demobilized by the SPLA/M (Coalition
2008).
In 2000 and 2001, the civil war was still ongoing, and prospects for
peace, or victory by either side, seemed bleak. The SPLA/M’s decision
to demobilize child soldiers, and thus weaken their own fighting capability is a puzzle in need of an answer.
The answer is that the SPLA/M was promised US financial and
moral support if they improved their human rights record. In May
2000, in a Senate Foreign Relations Committee hearing, Senator Sam
Brownback stated that he would work to get the US government to
provide Southern Sudanese opposition groups with non-lethal and
humanitarian aid, provided they “have developed procedures to
comply with verifiable international human rights standards” (US
Senate 2000, S Hrg 106–662, 46).7 However, Brownback’s statement
of support raises several important questions. Why would a US Senator
publicly voice his support for a rebel group with a very questionable
human rights record? And even if he did provide public support for the
SPLA/M, did the SPLA/M know that they had supporters in the US
7
The hearing was about the US Commission on International Religious Freedom’s
(CIRF) annual report, which listed a series of recommendations on how to
improve religious freedom in Sudan. One of the recommendations was to support
South Sudanese opposition forces in their fight against the Sudanese government;
it is this recommendation Brownback supports and publicly promises to address.
TANs, rebel groups, & demobilization of child soldiers
161
Congress? And did this somewhat vague statement of support lead the
SPLA/M to demobilize child soldiers? The answers to these questions
are all related to transnational mechanisms of diffusion; indeed, the
empirical evidence suggests that a boomerang effect – without a focus
on the naming and shaming of the perpetrator – was key.
The first observable implication of the boomerang model is the
presence of a domestic civil society that works to change the behavior
of human rights violators. These domestic groups also need to link up
with international actors to put pressure on the perpetrators from
above. The evidence clearly shows that this was the case in Southern
Sudan. Early on in the conflict, the SPLA/M worked hard to avoid the
establishment of any independent civil society organizations, domestic
or international. While there was a massive need for humanitarian
assistance in the mid to late 1980s the SPLA/M did not allow any
NGOs to establish bases in South Sudan (Rolandsen 2005, 30).
However, the SPLA/M’s view of civil society changed towards the
end of the 1980s and the early 1990s. In 1989, a large number of
churches formed the New Sudan Council of Churches (NSCC). This
group worked tirelessly for peace, justice, and human rights in Sudan
(Brown 2008). Even though the SPLA/M originally was against the
establishment of any independent civil society organization, they
accepted the NSCC for instrumental reasons – as a sign of tolerance,
to help them improve their standing internationally, and to increase
Western states’ willingness to provide aid (Rolandsen 2005, 76). The
key goal of the NSCC was not to weaken the SPLA/M; rather, they
worked against the “excesses and abuses conducted by the SPLA/M”
(Brown 2008, 205). However, the NSCC’s success was limited. Indeed,
when the SPLA/M split into several groups in 1991, human rights
abuses against the people of South Sudan reached new heights and
the NSCC began to look outwards for support and increased leverage.
In the US, they found a group of religious activists who were more than
willing to help put pressure on the US government to get involved in
the Sudan conflict.
NSCC’s effort to link up with US activists led to a broad coalition of
religious and anti-slavery human rights organizations. In turn, this
coalition sought to influence US policy-makers and members of
Congress, and pressured the US government to increase its involvement
to end the Sudanese civil war. The coalition viewed the war there
as between Islam and Christianity, as African versus Arab and as
162
Stephan Hamberg
slave-owning versus slaves; it saw the Sudanese government as the
perpetrator of grave human rights abuses, and almost uniformly supported the cause of the Southern Sudanese, if not the SPLA/M itself.8
Key individuals in the coalition and amongst the policy-makers kept in
close touch with the SPLA/M leadership, as well as with decisionmakers in the Clinton and later the Bush White House; they sought
to push President Bush into a policy of engagement and international
mediation between the government in Khartoum and the SPLA/M
(Hertzke 2004).
In retrospect, the shifts in US policy towards Sudan and the ongoing
civil war are both surprising and counterintuitive. Whereas the Clinton
administration had shown significant interest in Africa and peacemaking
efforts, President Bush was expected to concentrate on domestic politics
and certainly not to focus on the African continent. Indeed, the incoming
Bush administration had made it clear during the campaign that it was
reluctant to get involved in peacemaking or peacekeeping missions
(Woodward 2006, 113). However, in the context of Sudan, their roles
were reversed. The Clinton administration initiated sanctions against the
Khartoum government, froze diplomatic relations, and finally bombed
the Shifa pharmaceutical plant in August 1998. The Bush administration
would follow a very different and surprising strategy towards the conflict
in Sudan. Rather than isolating the government in Khartoum and continuing in the footsteps of Clinton, it engaged with both the government
and the SPLA/M. It also set in motion a protracted peace process that
ended in the Comprehensive Peace Agreement between the government in
Khartoum and the SPLA/M (Woodward 2006).
To understand why the Bush administration pursued peacemaking
in Sudan, we need to understand how the network of Christian human
rights organizations and key legislators came together to put pressure
on it.9 President Bush was influenced by religious groups prior to
taking office in 2001. Franklin Graham, a prominent Christian evangelical and the president and founder of Samaritan’s Purse, a Christian
relief and development organization with a long-standing commitment
to Southern Sudan, met with presidential candidate Bush two days
8
9
This is contrary to Uganda where, at least at the beginning of the insurgency, the
LRA was seen as the villain and the Ugandan government as the savior (Schmitz,
this volume).
In the remainder of the chapter, I refer to this group of organizations and
individuals as the US coalition.
TANs, rebel groups, & demobilization of child soldiers
163
prior to the 2000 election. In the meeting, Graham implored Bush to
help bring peace to Sudan. According to Samantha Power, Graham
told Bush, “[w]e have a crisis in Sudan. I have a hospital that has been
bombed. I hope that if you become President you’ll do something
about it” (quoted in Power 2004).
Though Graham was an influential actor, he was not alone. A wide
variety of religious groups, from all denominations had become interested in Sudan in the early to mid 1990s, and the groups were well
connected in Washington DC. They linked up with both Republican
and Democratic legislators and framed the debate about Sudan as one
about religious freedom and modern slavery (Hertzke 2004).10 Indeed,
they participated in congressional hearings and hosted members of the
NSCC, who themsleves participated in public events about the Sudanese civil war, met in private with key members of Congress, the State
Department and the National Security Council.11
A large number of legislators joined the cause, but three individuals
were particularly important: Senator Bill Frist, Senator Sam Brownback, and Congressman Frank Wolf.12 All three had been interested in
Sudan for several years prior to the election of Bush, but by combining
their efforts with the religious movement, and by appealing to President Bush’s ideas of compassionate conservatism, they now saw an
opportunity to make some headway on the issue.
Of the three senior politicians, Frank Wolf had the longest ties to
South Sudan. Wolf made his first trip there in 1989, where the SPLA/M
arranged a celebration for his entourage, which included other US
Congressmen; they left with newfound respect for John Garang, the
leader of the SPLA/M (Press 1989). Between 1989 and 2001, Wolf
traveled to Sudan 11 times, and was instrumental in passing the Sudan
Peace Act. However, Wolf was not the only one traveling to South
Sudan. Bill Frist traveled there several times in the late 1990s. In 1998,
he went with Franklin Graham to visit Samaritan’s Purse hospitals,
where Frist, a surgeon, performed surgery in a bush hospital bombed
10
11
12
For other examples of framing as a key transnational mechanism at work in civil
war, see the chapters by Bakke, Schmitz, and Adamson, all in this volume.
For example, the Institute on Religion and Democracy a Washington based
NGO hosted a number of NSCC members (personal communication, Faith
McDonnell, IRD, December 3, 2010).
These were the three most important individuals, but they got support from a
range of Democrats both in the House and the Senate (Hertzke 2004).
164
Stephan Hamberg
by government troops (McGrory 2001). During his time in Sudan, he
met John Garang several times, and subsequently called him a close
friend (Frist 2009). In his memoirs, Frist writes that though he did not
condone the behavior of the SPLA/M at all times, he certainly supported their cause (Frist 2009).
The story of Sam Brownback is not much different from that of the
two others. Brownback visited South Sudan for the first time in 1997,
and just as Frist, he called John Garang a friend (Brownback 2007).
Frist, Brownback, and Wolf, together with other legislators, policymakers, and religious organizations were instrumental in putting pressure on both the Clinton administration in its later years and the Bush
White House.
That the US coalition supported the cause of South Sudan and the
SPLA/M is clear, and it is reflected in several pieces of legislation. In
1999, the US Congress passed its first resolution in six years regarding
Sudan; it called for increased support for SPLA/M controlled territories, and condemned the government in Khartoum (Woodward 2006,
120–121). In December 2000, seven months after Brownback’s public
statement of support for the SPLA/M, and four months after its commitment to demobilize child soldiers, President Clinton signed legislation, sponsored by Senator Brownback, which allowed the US to
give humanitarian aid to the SPLA/M directly (Public Law 106–570,
Title V). In addition, the Sudan Peace Act passed the Senate in 1999,
but was not signed into law before 2002. This was perhaps a sign that
the coalition had more leverage on the Bush administration, which
assumed power in early 2001.
In addition to public statements and legislation passed to support the
SPLA/M, the Sudan Peace Act most clearly shows how the US coalition
supported South Sudan and the SPLA/M. The Act requested the president to take an active role in resolving the conflict between the government in Khartoum and the SPLA/M. Perhaps most significantly, it
stipulated a list of measures to punish the Sudanese government unless
it participated faithfully in future peace negotiations. However, if the
SPLA/M declined to participate or negotiate in good faith, there were
no repercussions, except for the fact that the Sudanese government
would not be penalized (Sudan Peace Act, 2002).
The evidence clearly shows that the NSCC worked to change the
SPLA/M’s behavior by linking up with international human rights
actors. It also shows that these actors put pressure on the Sudanese
TANs, rebel groups, & demobilization of child soldiers
165
government, through threats of sanctions and naming and shaming
(Wolf 2001; Shea 1998), while offering SPLA/M promises of support if
it improved its human rights record.
However, it is not clear whether US support for the South Sudanese
people and the SPLA/M influenced the group to demobilize its child
soldiers. Was the decision to demobilize a concession to religious
human rights organizations and US supporters, in return for US financial and moral support during future peace negotiations? Answering
this question requires that we show the SPLA/M was aware of the US
offer of support.
Again, the empirics indicate that the SPLA/M was well aware of
what was happening in the US at the time. For one, members of the
NSCC, who discussed these issues directly with US policy-makers, also
had direct contact with the leadership of the SPLA/M. According to
Brown, by the late 1990s, the leaders of the NSCC “were granted
personal access to Garang and his closest advisors and regularly utilized that ability to convey messages of personal accountability and
immediate implementation of human rights and good governance”
(Brown 2008, 210). In addition, the SPLA/M had representatives in
Washington who were kept informed about US efforts as well. Indeed,
according to Ted Dagne, a congressional staffer, the SPLA/M was even
queried on the drafting of some legislation.13
In addition to the NSCC leadership’s communication with the SPLA/
M, and the rebel group getting information directly from US policymakers, several members of the US coalition also remained in close
contact with the rebel movement. Deborah Fikes, the spokesperson for
human rights in the Midland Ministerial Alliance, kept in close touch
with John Garang (Brown 2008, 233). After the peace negotiations
began in Kenya in 2002, the US coalition sent a message to Garang
reminding him “that failure of the negotiations on grounds that leave
room to assign blame for the failure on the SPLM could sharply
undermine our capacity to maintain the support of the administration
and the United States Congress on your behalf” (quoted in Brown
2008, 234). This letter was sent long after the SPLA/M’s decision to
demobilize its child soldiers, yet it shows how transnational communication and pressure took place between the US religious coalition and
the SPLA/M.
13
Ted Dagne, personal communication, December 1, 2010.
166
Stephan Hamberg
In sum, the empirics demonstrate that transnational advocacy took
place between the NSCC, the US Church Alliance for Sudan, US
policy-makers, and the SPLA/M. For sure, there is no “smoking gun”
evidence that the SPLA/M’s decision to demobilize was a direct
response to US offers of support in exchange for an improved human
rights record. However, it is clear that the US offered support in return
for an improved human rights record, and that the SPLA/M knew
about this offer. In addition, we know that Senator Brownback stated
his support for such legislation on May 16, 2000, and that the SPLA/M
committed – in writing – to demobilizing its child soldiers in October
2000. Thus, the key observable implication for the concessions
mechanisms is satisfied, namely, that the offer was given prior to the
SPLA/M’s decision.
However, if the SPLA/M wanted to improve its human rights record,
why did it choose to focus on child soldiers? I argue this decision makes
sense for three reasons. First, it was something the SPLA/M leadership
could do as it did not require the rank and file to change their behavior.
As I discussed above, demobilization is an elite driven process. Even
without complete command and control of its forces, the leadership of
a rebel group can still make a decision to demobilize child soldiers and
then follow through on that commitment. This does not mean that
low-level officers will refrain from continuing to recruit or even rerecruit child soldiers; rather, it means the elites can control their
soldiers enough to collect the children and demobilize them. Second,
demobilizing child soldiers had symbolic value which could be verified
at the time it occurred. Indeed, when the SPLA/M demobilized the first
several thousand child soldiers, they did it in an official ceremony with
both UN officials and news media present (UNICEF 2001a).
Third, the SPLA/M leadership knew that child soldiers were being
frowned upon by the international human rights community. Increasingly during the 1990s, conventional human rights advocacy organizations had put a spotlight on child soldiers, by naming and shaming
rebel groups and governments that recruited them.
On the last point, however, we need to assess the degree to which
transnational advocacy networks specifically shamed the SPLA/M’s
use of child soldiers. Was the SPLA/M’s use of them singled out? Did
advocacy networks call for sanctioning the SPLA/M because of its use
of child soldiers? There is in fact very limited evidence that naming and
shaming played a direct role in the SPLA/M’s decision. This is not to
TANs, rebel groups, & demobilization of child soldiers
167
say that the SPLA/M was not shamed by international human rights
organizations. They clearly were. Yet, this shaming has been sporadic
and much of it came well before and after the SPLA/M’s decision to
demobilize its child soldiers.
The first report mentioning recruitment of child soldiers in Sudan was
by Amnesty International in 1991 (Amnesty International 1991). It
deals mainly with human rights abuses by the Sudanese government;
human rights violations by the SPLA/M are given one page, and its
recruitment of child soldiers, one sentence. In subsequent reports,
Amnesty rarely discussed the SPLA/M’s use of child soldiers. Human
Rights Watch (HRW) has been more active in reporting on the SPLA/M’s
use of child soldiers. Out of 17 reports on Sudan from 1990 to 2001,
the issue of child soldiers and the SPLA/M was mentioned in nine of
them. Some of these reports were dedicated almost entirely to the
issue of the SPLA/M and child soldiers (Human Rights Watch 1994,
1996). However, most were about human rights violations in general
in Sudan; these would often include sections on the SPLA/M and
child soldiers.
There was thus clearly some naming and shaming of the SPLA/M
and its use of child soldiers in the mid 1990s, but other rebel groups
have received much more attention in this regard.14 In 1998, five
international NGOs created the Coalition to Stop the Use of Child
Soldiers, and much of the naming and shaming of states and rebel
groups has since taken place through it. In the beginning, the Coalition
sent out press releases and newsletters, and in 2001 it published the
first of three global reports on the use and recruitment of child soldiers.
In this first report, the SPLA/M was praised for demobilizing over
3,000 child soldiers in 2001, but was also criticized for having more
than 7,000 child soldiers remaining in its ranks (Coalition to Stop the
Use of Child Soldiers 2001).
Human rights organizations and the Coalition also pushed the UN to
take a stronger stand on the issue. In 1997, the UN General Assembly
passed a resolution which requested that the Special Representative of
the Secretary General for Children and Armed Conflict submit annual
reports to the General Assembly on how children are affected by armed
14
Especially the Lord’s Resistance Army in Uganda, the Revolutionary United
Front in Sierra Leone and the LTTE in Sri Lanka. However, see Schmitz, this
volume, for an assessment of how little shaming influenced the LRA’s behavior.
168
Stephan Hamberg
conflict around the world (UN General Assembly 1997). In 1999, the
Security Council passed resolution 1261, where it condemned the use
and recruitment of child soldiers. The resolution urged relevant parties
to demobilize them, and further urged UN agencies to make sure their
personnel were trained and equipped to deal with war-affected children. Finally, the resolution requested the Secretary General to submit
a report by July 31, 2000, on the implementation of the resolution.
In addition to Resolution 1261, the Security Council has passed five
resolutions specifically dealing with the issue of child soldiers (UN
Security Council 2000, 2001, 2003, 2004, and 2005). The most important of these are resolutions 1379 and 1460, which request the Secretary
General to submit a report to the Security Council detailing which
groups and states use and recruit child soldiers and to what degree these
groups have made progress in demobilizing them. Looking at the timing
of these resolutions, most were passed after the SPLA/M made its decision to demobilize child soldiers; it thus seems unlikely that any mentioning of the SPLA/M in reports from the Secretary General influenced
its 2000–2001 decision to demobilize children.
Examining the UN’s effort to name and shame the SPLA/M prior to
2000 reveals a rather weak effort. UN General Assembly Resolutions
and notes by the Secretary General on the situation of human rights
in the Sudan from 1993–2001 indicate that the UN largely focused
on shaming the Sudanese government and only sporadically mentioned
the SPLA/M by name regarding its use of child soldiers.15 The
first report on the human rights situation in Sudan by the Secretary
General’s Special Rapporteur was written in 1993 (UN 1993). It
focused heavily on human rights abuses by the government of Sudan,
but a brief section included SPLA/M abuses, including recruitment of
child soldiers. Later reports by the Secretary General on the human
rights situation in Sudan followed a similar pattern. In all the
reports, the main focus is on human rights abuses by the Sudanese
government, and only minor sections are related to the recruitment of
child soldiers.
When recruitment of child soldiers is mentioned, it is almost always
discussed as a problem relating to all parties in the conflict. For
example, in the 1994 report, the Special Rapporteur wrote that:
15
In 1993, the UN Commission on Human Rights asked the Secretary General to
submit annual reports on the human rights situation in Sudan.
TANs, rebel groups, & demobilization of child soldiers
169
based upon his own findings and on information from several reliable
sources, the Special Rapporteur considers the problem of unaccompanied
minors and the use of children as soldiers by all parties to still be of great
concern, despite repeated calls from the international community to put an
end to this practice. (UN 1994, 14)
In terms of UN General Assembly Resolutions, the situation is very
much the same. In all of the resolutions, the Sudanese government
bears the brunt of the critique, and when the resolutions specifically
deal with child soldiers the language is almost identical to that of the
Secretary General’s reports. As this shows, the shaming of the SPLA/M
and its use of child soldiers by the UN has been weak and sporadic; the
UN has mainly shamed the Sudanese government regarding its human
rights record.
In sum, it is clear that the human rights advocacy networks successfully
brought the issue of child soldiering to people’s attention, and their work
probably led the UN Security Council to pass several resolutions in the
early to mid 2000s. However, neither the advocacy networks nor the UN
consistently shamed the SPLA/M’s use of child soldiers prior to its decision to demobilize them. Indeed, my analysis indicates that naming and
shaming of the SPLA/M by mainstream human rights organizations
played little or no direct role in its decision. In addition to the lack of
consistent shaming by traditional human rights NGOs, South Sudan’s
supporters in the US never publicly named and shamed the SPLA/M for
its human rights abuses, including the use of child soldiers.
Rather than name and shame the SPLA/M, the US coalition concentrated its energy on shaming the Khartoum government. Both publicly
and behind closed doors, they focused on portraying the Sudanese
government as the perpetrator of grave human rights abuses and the
people of South Sudan as victims (Wolf 2001; Shea 1998). Indeed, they
rarely publicly supported the SPLA/M. Instead they “focused on how to
frame arguments to avoid getting bogged down in side battles that were
hard to understand or justify, such as the nature of the SPLA/M”
(Hertzke 2004, 274). However, the underlying sentiment was one of
support and admiration for the SPLA/M and especially of John Garang
(Belz 2001). Consider the following statement by one of the founders of
the Church Alliance for Sudan, a key organization in the US coalition:
I was/am on the side of the SPLA/M, and loved and respected John Garang
very much, as I do Salva Kiir, as well. I believe that they were fighting a
170
Stephan Hamberg
righteous, just war to protect and defend the people of South Sudan and
Nuba Mountains from jihad and the imposition of Shariah, annihilation of
the black, African people of Sudan.16
The point was not that the US coalition was unaware of SPLA/M’s
human rights abuses; rather, they believed the Sudanese government’s
abuses were so much worse, that they focused exclusively on those.17
In a manner similar to the shaming argument, both the domestic
level approaches discussed above also fail to explain the SPLA/M’s
decision to demobilize its child soldiers. In the period 1999–2001, the
dynamics of the war are best characterized as a stalemate and the
SPLA/M’s financial situation had deteriorated rather than improved
in the years prior to its decision.
The core claim with the battlefield dynamics argument is that insurgencies will demobilize child soldiers if they are winning the war. If this
argument explains the SPLA/M’s behavior, we would expect it to be
doing well on the battlefield and to publicize its decision to demobilize
child soldiers internationally and domestically. The situation on the
ground in Sudan in 2000–2001 does not support this argument. By the
early 2000s, the best description of the battlefield dynamics was a
military stalemate, with the government having a somewhat stronger
hand. In other words, the SPLA/M was neither winning nor losing the
war when it made its decision to demobilize child soldiers. In addition,
there is no evidence that the SPLA/M tried to publicize its decision to
demobilize them. Except UNICEF press releases and NGO reports,
I have not uncovered any news articles about its decision.
The second domestic explanation considers the financial strength of
the rebel group. One would expect a rebel group to demobilize (relatively cheap) child soldiers when and if its financial situation improves,
as this allows the group to recruit more efficient soldiers, even if they
cost more. This was clearly not the case for the SPLA/M.
For the first 15 years of the conflict (1983–1998), the SPLA/M was
financially supported by several of Sudan’s neighbors. It maintained
bases in Ethiopia and received support from both Eritrea and Uganda.
According to Woodward, Eritrea, Ethiopia, and Uganda “felt
16
17
Personal communication, August 27, 2010.
Roger Winter, personal communication November 30, 2010. Winter is a former
State Department and USAID employee. For more on him and his sympathies
for the SPLA/M, see Griswold 2008.
TANs, rebel groups, & demobilization of child soldiers
171
threatened by the subversive activities of Islamists that they saw as
sponsored by Sudan,” and they were in a position to actively intervene
(Woodward 2006, 94). All three countries were also encouraged by the
US to support the SPLA/M in its fight against the regime in Khartoum.
In 1996, the US sent $20 million in military aid to Eritrea, Ethiopia,
and Uganda, and according to Woodward, it was expected that some
of this money would end up with the SPLA/M (Woodward 2006, 98).
This support helped the SPLA/M improve its military position.
Through major offensives in 1997–1998, people began to believe that
the regime in Khartoum was nearing its end – but the SPLA/M did not
manage to win the civil war. These hopes were further dashed when
the war between Eritrea and Ethiopia broke out in 1998. At the same,
the Ugandan government became increasingly involved with its own
war against the LRA (Woodward 2006, 99). While the SPLA/M was
losing financial and material support from outside states, the Sudanese
government was able to increase its revenues through oil exploration.
This improved its ability to fight the SPLA/M, while the latter retreated
from places it had taken just years earlier.
Therefore, when the SPLA/M announced its decision to demobilize
child soldiers in 2000–2001, it was much worse off financially than in
1997, and accordingly we would expect it to recruit child soldiers, not
demobilize them. There is also no evidence that it recruited adult
soldiers to substitute for the demobilized child ones.
Conclusion
The SPLA/M’s decision to demobilize its child soldiers in 2001 was a
major milestone. No armed group has demobilized the majority of its
child soldiers while still fighting a civil war. The empirical analysis
shows that transnational advocacy by US and Southern Sudanese civil
society actors towards US politicians and the SPLA/M led the rebel
movement to demobilize child soldiers to secure financial and political
support from the United States. Southern Sudanese church organizations linked up with US religious organizations, which in turn put
pressure on key US policy-makers to not only get involved in the
Sudanese civil war, but to explicitly support the SPLA/M. Rather than
a traditional boomerang pattern, where domestic groups seek international support to name and shame human rights violators, this case
shows that the New Sudan Council of Churches pressured the SPLA/M
172
Stephan Hamberg
from below, while simultaneously securing support for the rebel group
from the US government.
The fact that the SPLA/M demobilized child soldiers to secure US
support speaks to the importance that positive incentives played in its
decision. Even though it is difficult to draw generalizations from this
one case study, my findings support Katzenstein and Snyder’s (2009)
argument that human rights organizations need to pay more attention
to the incentives that rights violators face. Naming and shaming can
impose significant costs on actors, but for rebel groups or other nonstate armed groups, these costs rarely outweigh the benefits of future
violations.
|
7
Conflict diffusion via social identities:
entrepreneurship and adaptation*
martin austvoll nome and
nils b. weidmann
The grand narratives of armed civil conflicts – through which participants
explain their actions and observers identify the clashing interests – link
them to social identities. In El Salvador, 1992, insurgents invoked the
distinction between the propertied and the landless (Wood 2003, 1–2); in
Cyprus, 1974, with a Turkish intervention looming, the leader of the
Turkish-Cypriot community emphasized the indivisibility of Turkishness
in Turkey and Cyprus (Markides 1977, 33); and in Chechnya, 2007, one
insurgent leader described as paramount the opposing identities of
believers and unbelievers (Bakke, this volume).
In the effort to understand the mechanisms that lead armed civil
conflicts to diffuse from one country to another, it is therefore appropriate to give attention to conflict diffusion via social identities. How might
particular social identities diffuse across international boundaries?
Social identities are social categories that can be characterized by
several sorts of content. Abdelal et al. (2009, 19) argue that social
identities may contain sets of goals shared by members of the group,
particular views of other social identities, and worldviews or understandings of conditions and interests. We focus on a fourth type of
identity content: its constitutive norms. Abdelal et al. (2009, 19)
describe these as the formal and informal rules that define group
membership. In other words, the constitutive norms of a social identity
are its boundary-drawing properties. By defining the boundaries of
social identities, constitutive norms should be of particular interest
for students of conflict. By modeling the transnational diffusion of
norms, we seek to capture one aspect of the diffusion of armed civil
conflicts.
* For valuable comments on earlier versions of this chapter, we thank Andy
Bennett, Jeff Checkel, Kristian Skrede Gleditsch, Matthew Hoffmann, Jay Lyall,
Idean Salehyan, and the members of the Working Group on Transnational and
International Facets of Civil War, Centre for the Study of Civil War, Peace
Research Institute Oslo.
173
174
Martin Austvoll Nome & Nils B. Weidmann
Diffusion is notoriously difficult to observe. The diffusion of conflicts,
as the diffusion of other phenomena, is a process, not an outcome (Elkins
and Simmons 2005). Armed civil conflicts have diffused, rather than
merely occurred in sequence, when the onset of one conflict is a consequence of processes set in motion by conflicts elsewhere. As Checkel
argues in his introduction to this volume, the emphasis on process suggests that diffusion is a large class of mechanisms linking one civil conflict
with another. It is understandable then that the comparativist literature
on conflict diffusion struggles with good measures, often settling for
measures of its outcome, or for proxies of scope conditions. Diffusion is
assumed.
In line with this book’s ambition to examine transnational conflict
processes, we offer an alternative way to explore mechanisms of diffusion, using agent-based modeling. This technique allows us to posit
that particular mechanisms of diffusion are at work, to model them
closely, and to assess the plausibility of such mechanisms by using
computer simulation. If the results of these simulations resemble empirical patterns of conflict diffusion, then the underlying mechanisms are
candidate explanations of diffusion.
As we develop our model, we begin from the understanding that
conflict diffusion via social identities is a transnational process. The
importance of transnational factors in civil wars is well established
(Checkel, this volume). At the same time, a strong research tradition on
transnationalism in international relations (IR) theory could contribute
much to conflict research. We thus turn to this transnationalist literature,
positing two alternative mechanisms of diffusion and compare how they
perform. One mechanism involves norm-driven individuals who seek to
behave appropriately in what they define as their social context. We
observe the populations in two separate countries and the effect of
transnationalizing the social context in which individuals interact. We
call this mechanism “social adaptation.” The other mechanism involves
single individuals – norm entrepreneurs – who have the power to alter
people’s repertoires of imaginable behavior. Via this mechanism, norm
entrepreneurs affect the latent norms in a society. We call this mechanism
“norm entrepreneurship.”
By using the technique of agent-based modeling, we are able to model
social adaptation and norm entrepreneurship directly, and assess their
consequences in an environment of norm-driven individuals. Rather than
assuming their operation when observing the outcome of diffusion or its
Conflict diffusion via social identities
175
scope conditions, we know that these mechanisms operate at the outset,
and then explore their effects. From a modeling perspective, our objective
is simple. We seek to create an abstract heuristic model designed to
explore some fundamental logics of diffusion. Our aim is to develop a
model that generates macro-patterns that qualitatively match empirical
patterns uncovered in this volume (and elsewhere). Anecdotal evidence
suggests that both social adaptation and norm entrepreneurship are
plausible mechanisms of diffusion.
One such anecdote leads into the remainder of this chapter: We
begin the next section by briefly describing the spread of Kurdish
nationalist contention from Turkey to Germany to illustrate the sort
of diffusion we explore. In the same section, we review the mechanisms
of social adaptation and norm entrepreneurship. The subsequent
section makes the case for agent-based modeling, explains its basics
and logic of inference. The final two sections present our model in full.
First, we model the emergence of a norm in a single country. Second,
we link two such countries: one source country from which norms can
diffuse, and one recipient country that might be affected by developments in the source country. We present three scenarios: norm emergence in two closed polities, diffusion by transnationalized social
adaptation, and diffusion by transnationalized norm entrepreneurship.
We then discuss our observations by linking them to evidence from
other chapters in the book, and by suggesting ways in which our model
might add to the understanding of their empirics. We conclude by
noting that norm entrepreneurship seems to be the more robust mechanism of diffusion, and by pointing to possible model extensions.
Conflict diffusion as identity diffusion
As an example of the phenomenon we explore, consider the immigrant
communities in Germany with origins in Turkey. Up until 1996,
Germany became a new arena of sometimes violent activism in the
armed civil conflict between the Workers’ Party of Kurdistan (PKK)
and the Turkish state. This diffusion of conflict manifested itself in
events such as large-scale demonstrations, hunger strikes, the targeting
of Turkish-owned businesses for attack, and threats of violence against
Germans in efforts to pressure the German government to promote the
PKK’s cause (Lyon and Ucarer 2001). Prior to this escalation of
activism, the immigrant community from Turkey experienced a change
176
Martin Austvoll Nome & Nils B. Weidmann
in social identities in which many immigrants with some Kurdish
ancestry began to view themselves as Kurdish and sympathize with
Kurdish separatism. Lyon and Ucarer (2001, 932–933) characterize
this change as a “development of the collective insurgent consciousness.” This outcome was a consequence of the diffusion of social
identity, including its constitutive norms. It involved a process where
the social context of the PKK – Turkey conflict was transnationalized,
and in which norm entrepreneurs were active.
By the time the recruitment of Turkish guest workers was halted
in 1973, about 900,000 Turkish citizens had settled in Germany
(Leggewie 1996, 81). Although many of them had some Kurdish ancestry, only a few called themselves Kurds (Leggewie 1996, 81). In a change
away from economic migration, Kurds began in the 1980s to apply for
political asylum in Germany. Their reasons for migrating were political
to a larger extent than before, and among the immigrants were dedicated supporters of Kurdish separatism, as well as members of the PKK
(Leggewie 1996, 81; Lyon and Ucarer 2001, 931). The political liberties
in Germany allowed immigrants to discuss the Kurdish question openly
(Leggewie 1996, 82); it was now “possible to explore and express
Kurdish cultural and linguistic identity” (Lyon and Ucarer 2001, 933).
Kurdish cultural associations became arenas where a transnationalized
social context could manifest itself in the consolidation of Kurdish
identity (Lyon and Ucarer 2001, 933). Research on the socialization of
new group members suggests that individuals are more likely to adopt
in-group norms when they have few prior ingrained beliefs, and when
they interact with authoritative members of the in-group (Checkel
2001). When economic migrants of Kurdish origin were exposed to
Kurdish nationalism in such small-group settings, one might imagine
that they were likely objects of socialization into the norms of Kurdish
nationalist identity. The norm entrepreneurs in this process were
the PKK, who infiltrated cultural associations to spread their version
of Kurdish nationalism (Leggewie 1996; Lyon and Ucarer 2001).
In sum, the diffusion of social identities from Turkey to Germany was
a consequence of the transnationalization of the social context of the
PKK – Turkish conflict and of the activities of PKK norm entrepreneurs.
In its wake, Germany experienced sometimes violent events in the
conflict between the PKK and Turkey.
The diffusion of Kurdish nationalist activism from Turkey to Germany
raises not only the question of diffusion mechanisms; the case also raises
Conflict diffusion via social identities
177
the question of how social identities and armed civil conflicts are linked.
By modeling the diffusion of social identities’ constitutive norms, we
model the diffusion of rules that determine group membership. In other
words, we model the diffusion of social identities’ boundary-drawing
properties. Work on how the boundaries of social identities relate to
violent civil conflict is too extensive to review in full here. However, we
need to make clear some basic assumptions.
Had violence in civil conflicts followed the fault lines of readily
observable social identities, then the task for social scientists would
have been an easy one. Our model depends not on a one-to-one
association between social identities and the behavior of perpetrators
and targets in civil wars. Rather, we assume that social identities
provide the grand narratives within which violence unfolds and is
perpetuated. These are narratives that elites, perpetrators, and victims
take part in prior to, during, and after violence. By linking violence to
social identities, both perpetrators and victims take part in the coding
of armed civil conflicts. One must expect actors in civil wars to be
influenced by prevailing interpretive frames that in turn are supplied by
changing social identities (Brubaker and Laitin 1998, 428; see also the
discussion of framing in Bakke, this volume). We follow Brubaker and
Laitin (1998) when they argue that conflict and violence should be
analyzed as separate phenomena.
That is not to say that social cleavages and violence are unrelated. By
combining the concepts of cleavage and alliance, Kalyvas (2003) is able
to link conflict and violence. To Kalyvas, a cleavage is “a symbolic
formation that simplifies, streamlines, and incorporates a bewildering
variety of local conflicts” (Kalyvas 2003, 486); it is the structure that
links actors at the center with action on the ground. Actors at the
center provide resources and a context of impunity that allow actors
in the periphery to let a variety of local cleavages unfold in violence.
Center and periphery are linked by alliance. As Kalyvas puts it “[the
concept of] alliance allows us to see civil wars as concatenations of
multiple and often disparate local cleavages, more or less loosely
arrayed around the master cleavage” (Kalyvas 2003, 486). The constitutive norms that define the boundaries of social identities provide the
raw material for such master cleavages. As such, particular social
identities come close to being a necessary condition for armed civil
conflict. Any account of the diffusion of civil conflicts must therefore
include an account of the diffusion of social identities.
178
Martin Austvoll Nome & Nils B. Weidmann
Social adaptation and norm entrepreneurship
Social adaptation and norm entrepreneurship are our two candidate
mechanisms of conflict diffusion. Recalling Checkel’s brief survey of
the literature on causal mechanisms (this volume), social adaptation
and norm entrepreneurship are pathways or processes by which the
diffusion of social identities is produced (see also Gerring 2007b, 178);
they are “relational and processual concepts . . . not reducible to an
intervening variable” (Falleti and Lynch 2009, 1149). By using agentbased modeling, we are able to closely approximate these relational
and processual phenomena.
Social adaptation and norm entrepreneurship are by no means the
only mechanisms of conflict diffusion. Bennett’s (this volume) taxonomy of mechanisms is helpful in assessing the ground we do and
do not cover. Bennett classifies social mechanisms along two dimensions that generate a reasonably exhaustive taxonomy. One dimension
distinguishes mechanisms that are rooted in power, functional efficiency, and legitimacy. By focusing on the constitutive norms of social
identities, we limit the paper at the outset by not considering explanations based on power or functional efficiency. Since norms are standards of appropriate behavior, they evoke notions of legitimacy. In
Bennett’s 12-quadrant property space it is only the top row that we
are considering.
A second dimension of the taxonomy classifies mechanisms by their
direction of interaction between structures and agents: agents interacting with agents, structures affecting agents, agents affecting structures,
or structures relating to structures. As we conceive of them, social
adaptation is a structure-to-agent mechanism, and norm entrepreneurship is an agent-to-structure mechanism.
Social adaptation is a mechanism by which individuals alter their
behavior to conform with the norms of their social context. People are
norm-driven. Individuals change their conduct once they have
developed a notion of the standards of behavior in their society. As
such it is a structure-to-agent mechanism. A society develops an intersubjective understanding of behaviors that are more and less desirable,
and in turn, individuals adapt to these understandings. As Bennett (this
volume) puts it in his taxonomy, norms work as an enabler and
constraint on action. Norms that are constitutive of social identities
are in this manner internalized or habituated in a process where norms
Conflict diffusion via social identities
179
bias choice. In the words of Abdelal et al. (2009, 21), “certain behaviors are consciously ruled out or discounted as inappropriate for one’s
identity.”
By labeling this mechanism “adaptation,” we wish to distinguish it
from related mechanisms such as mimicking and emulation. Mimicking is the mechanism by which a novice copies the behavioral norms of
a group as a reaction to uncertainty in situations where the costs of not
fitting in are high (Johnston 2008, 45). In contrast, social adaptation is
a mechanism not reserved for novices. We expect individuals that are
well established in a society to continuously try to fit in. Emulation, in
turn, is the “conscious and careful search for exemplars and success
stories, a dissection of the reasons for their success, and the application
of these lessons to the maximization of some specific expected utility”
(Johnston 2008, 45–46). Where emulation is a search for methods,
procedures, or tactics, social adaptation is a move towards
appropriateness; where emulation emphasizes highly rational maximizing behavior, social adaptation involves less utility-maximizing
processes of internalization and habituation.
In dictionary usage, adaptation is described as “adjustment to environmental conditions,” or the “modification of an organism or its parts
that makes it more fit for existence under the conditions of its environment” (Merriam-Webster), and as “the process of changing to suit
different conditions” (Cambridge).1 In ecology, adaptation is understood as “the adjustment or changes in behavior, physiology, and
structure of an organism to become more suited to an environment.”2
In an analogous social environment, this is precisely the mechanism we
seek to capture. Might social adaptation be a mechanism of the diffusion of social identities?
We ask the same question of norm entrepreneurship. Norm entrepreneurship is a mechanism by which individuals who wish to influence social identities, and who have some platform from which to
propagate their views, attempt to affect the norms in a society. Norm
entrepreneurship is an agent-to-structure mechanism. Individual
agents, often with an institutional platform, attempt to influence the
1
2
www.merriam webster.com/dictionary/adaptation (accessed July 31, 2010);
http://dictionary.cambridge.org/dictionary/british/adaptation (accessed July 31,
2010).
www.biology online.org/dictionary/Adaptation (accessed July 31, 2010).
180
Martin Austvoll Nome & Nils B. Weidmann
standards of appropriate behavior that are held intersubjectively in a
population.
One often associates norm entrepreneurs with Finnemore and
Sikkink’s (1998) work on norm emergence in the society of states.
To Finnemore and Sikkink, norm entrepreneurs’ methods of choice
are persuasion and framing; they “attempt to convince a critical mass
of states . . . to embrace new norms” (Finnemore and Sikkink 1998,
895; see also Payne 2001, 38–39). Many scholars of transnationalism
have found it useful to classify certain actors as norm entrepreneurs.
Recent examples include Williams (P. Williams 2007, 255), who
traces how several actors have engaged with the African Union to
affect its range of legitimate policy options; Capie (2008), who
recounts how international bureaucracies such as the United
Nations Department of Disarmament Affairs and single states such
as Canada and Japan have acted as norm entrepreneurs to affect the
Association of Southeast Asian Nations’ policies on small arms;
Rushton (2008), who describes Boutros Boutros-Ghali as an erstwhile
norm entrepreneur in his efforts to promote democratic governance
when serving as General Secretary of the United Nations; Mantilla
(2009), who analyzes efforts to affect the human rights practices of
transnational corporations; and Saurugger (2010), whose work is on
the participation of civil society in EU decision-making processes.
This work may be varied, but the listing of recent applications
of the norm entrepreneur concept suggests they are safely situated
within the realms of international and transnational organizations.
By modeling the activities of norm entrepreneurs in the context of
conflict diffusion, we transpose norm entrepreneurship to the field
of conflict research.
One might reasonably ask who these norm entrepreneurs are.
Finnemore and Sikkink (1998, 896) describe them as “agents having
strong notions about appropriate or desirable behavior in their community,” whose motivations are “difficult to explain . . . without reference to empathy, altruism, and ideational commitment” (898). A close
analogy to norm entrepreneurs in the literature on conflict diffusion is
Kuran’s (1998) “ethnic activists,” who are “self-motivated to promote
public ethnic behavior among co-ethnics.”
Where Finnemore and Sikkink’s norm entrepreneurs target a society
of states with a limited number of members, our norm entrepreneurs
operate in societies of individuals whose numbers are much greater.
Conflict diffusion via social identities
181
They have less of an opportunity to engage directly with people, and
are therefore less able to use persuasion in an interactive process.
Instead, our norm entrepreneurs have a one-directional impact. By
propagating their views about social identities, they affect people’s
possibilities of action by altering the repertoire of imaginable behavior
for individuals in the target audience. Norm entrepreneurs do not
affect behavior directly. Rather, they introduce latent behavioral
norms. By reaching some or all individuals in the target audience
simultaneously, norm entrepreneurs create latent possibilities of action
that are present in the entire group of receptive individuals. These
possibilities of action only manifest themselves in people’s behavior
when they begin to observe an emerging norm. Through norm entrepreneurship, the mass of individuals is thus changed, but the change
only manifests itself in behavior if people deem the new standard to be
the most appropriate one. Therefore, in order to result in changed
behavior, norm entrepreneurship depends on subsequent social
adaptation.
We consider neither structure-to-structure nor agent-to-agent mechanisms. The former is most easily justified. The method of agent-based
modeling precludes structure-to-structure mechanisms because it
depends on specifying interaction on the micro level. True, agentbased models can link structural developments. These are the “emergent phenomena” in modeling language. However, agent-based
models can only link structural developments via mechanisms that
connect individual agents. Those with a commitment to methodological individualism would approve (see for example Coleman 1986;
Hedström and Swedberg 1998). Rooted in Coleman’s macro-micromacro model of collective social action, Hedström and Swedberg
(1998, 21–22) argue that social phenomena on the macro level can
only be linked via three sorts of mechanisms: situational mechanisms
through which a macro state influences individual actors; actionformation mechanisms that relate individuals to each other; and transformational mechanisms through which individuals’ actions generate
new macro states. To Coleman and Hedström and Swedberg there are
no structure-to-structure mechanisms. Whether one agrees with them
or not, it remains true that agent-based modeling is only compatible
with notions of social life in which individual actors hold prominence.
The fact that we leave out agent-to-agent mechanisms has less of
a methodological rationale. Agent-to-agent mechanisms such as
182
Martin Austvoll Nome & Nils B. Weidmann
persuasion, socialization, complex learning, and coercion, are probably important micro processes when social identities spread from one
country to another. They are mechanisms that should be explored in
agent-based models of norm diffusion. However, we exclude them
from the present model primarily to retain its simplicity and replicability. By starting with the most simple model and then adding features
step by step, it allows us to incrementally make it more complex and
not lose sight of its foundations. In this chapter, we add a second
country to a base model of only one country. This move allows us to
introduce two features. We transnationalize the social context in which
social adaptation occurs, and we turn norm entrepreneurs into transnational agents of norm diffusion. In order to improve the explanatory
value of the model, there is little doubt that we should add agent-toagent mechanisms in future work.
The case for agent-based modeling
In a volume that arguably privileges empirical work, how might agentbased modeling contribute to an understanding of conflict diffusion?
Agent-based modeling has advantages because diffusion can be difficult to observe, and because it forces the modeler to carefully specify
and operationalize possible mechanisms of diffusion. Mechanisms are
difficult to observe by any standard. Indeed, empirical researchers do
not observe mechanisms directly, but at best their observable implications (Checkel, this volume; Hamberg, this volume, Bennett, this
volume).
Some work on conflict infers diffusion from rather indirect proxies.
Its focus is not on diffusion mechanisms as such, but their conditions,
typically means of communication and determinants of receptivity.
Thus Hill and Rothchild (1986) observe that domestic conflict is more
likely when there is a certain level of conflict in the region (from which
conflict can diffuse) and when there is a greater level of ethno-linguistic
fractionalization (meant to suggest a latent sense of group identity),
when the per capita number of radios is greater (indicating means of
communication as a condition for diffusion), and when the country has
experienced earlier domestic conflict (designed to indicate receptivity).
Buhaug and Gleditsch (2008) observe that domestic conflict is more
likely when there is an ongoing domestic conflict in a contiguous
country, suggesting that it proxies a condition for diffusion because
Conflict diffusion via social identities
183
reference examples and media attention primarily focus on events in
nearby states (Buhaug and Gleditsch 2008, 220). Finally, both Buhaug
and Gleditsch and Forsberg (2008) find that domestic conflict is more
likely when groups can be ascribed the same ethnic identity as crossboundary groups in a conflict zone. Transnational ethnic ties would
seem to be understood as a scope condition of diffusion, not an
observable implication of any diffusion mechanism as such.
In short, the difficulties of observing mechanisms of diffusion lead
some to measure conditions for diffusion, and not any observable
implications of diffusion mechanisms. Agent-based modeling enjoys
an advantage in this regard because it allows us to operationalize and
measure specific mechanisms in a virtual environment.
By requiring the modeler to specify how individual agents interact,
agent-based modeling has the added advantage of inducing the explicit
theorization of diffusion mechanisms. After beginning deductively and
developing an informal model of the workings of social adaptation and
norm entrepreneurship as mechanisms of diffusion, we then develop an
agent-based model by proceeding in four steps (Hoffmann 2008, 188–
189). In general terms, we first create artificial agents – individuals in
two virtual countries – endowing them with characteristics such as
particular behavioral rules, the ability to perceive their environment,
and the ability to take decisions and act on them. These characteristics
are simply a conjecture about how social adaptation and norm entrepreneurship work on the micro level to generate the foundations of
conflict on the macro level. We then define the social environment in
which our agents interact. Third, we simulate histories in this world,
and record these histories in real time – a process analogous to retrospectively gathering data from real-world histories. Finally, we analyze
these histories by graphic means and statistical comparisons.
Generally speaking, the purpose of agent-based modeling is to
understand macro-level patterns or processes by locating their generative social mechanisms (Cederman 2001, 17). In the metaphorical
language of Epstein and Axtell, agent-based simulations are like
“laboratories, where we attempt to ‘grow’ certain social structures in
the computer – or in silico – the aim being to discover fundamental
local or micro mechanisms that are sufficient to generate the macroscopic social structures and collective behaviors of interest” (Epstein
and Axtell 1996, 4, emphasis in the original). Agent-based modeling
looks at first glance like a purely quantitative method, since the agents’
184
Martin Austvoll Nome & Nils B. Weidmann
characteristics and behavioral rules are represented by numbers.
However, as Hoffmann (2008, 188) emphasizes, agent-based modeling
is at least as much a qualitative method. Its lack of reliance on closedform analytical solutions and its possibility of generating many different simulation histories render it complementary to game-theoretic
approaches and large-N statistical analyses. Thus, it is much better
suited to represent notoriously fuzzy concepts such as emotions or
norms, which have traditionally been the domain of qualitative
researchers. By adding a level of higher theoretical precision to purely
verbal accounts, the simulation allows us to examine more closely the
implications of a theoretical framework.
The inferences one might draw from such a procedure are modest,
but non-trivial. By beginning with an informal model of the workings
of social adaptation and norm entrepreneurship, agent-based modeling
offers a means to assess the logic of that verbal model (Hoffmann
2008, 196). Should we be able to grow the diffusion of social identities
from our conjectures about the workings of the mechanisms, then we
would have a “candidate explanation” of diffusion. Such candidate
explanations could be a useful theoretical starting point for subsequent
empirical research.
A model of norm emergence
We present the model of the diffusion of social identities in two steps.
In this section, we first describe how we model the mechanisms of
social adaptation and norm entrepreneurship within one country.
Neither mechanism is necessarily transnational, so for clarity we first
describe how they work in a “closed polity” setting (Gleditsch 2007).
In the next section, we then simulate how social adaptation and norm
entrepreneurship might work as mechanisms of diffusion by transnationalizing the context in which they operate. It is by adding a second
country and linking the two societies with our candidate mechanisms
that we can assess how social adaptation and norm entrepreneurship
might work to diffuse social identities.
To model the emergence of the constitutive norms of social identities
in a single society, we need a social environment in which agents are
norm-driven. Hoffmann (2008) provides such an environment. We
adopt his model of norm emergence in single societies as a foundation
for developing our model of diffusion. Hoffmann applies his model to
Conflict diffusion via social identities
185
exploring Finnemore and Sikkink’s (1998) theory of the norm life cycle
among states in the international system. In other words, Hoffmann
conceives of his agents as countries. In contrast, we think of the agents
as individuals. Let us describe how we replicate Hoffmann’s model for
any one country.
The essential conjecture of the model is that agents are norm-driven.
Via social adaptation, agents seek to behave appropriately according
to a standard that is set intersubjectively, that is, by the collective
behavior of their society at any one time. In our computer simulation,
agents make their behavior known to all other agents by picking a
number between 0 and 100. This is not a random pick. Agents are
limited at any one time by an individual repertoire of imaginable
behavior. This repertoire is operationalized as a set of three norms,
or rules. Representing the repertoire by exactly three rules is a modeling choice we adopt from Hoffmann (2008) – there is no theoretical
reason to use this number, and others are possible. These three rules
come from a universe of seven rules, where rule 0 limits behavior to the
range 0–10, rule 1 limits behavior to 15–25, rule 2: 30–40, rule 3: 45–
55, rule 4: 60–70, rule 5: 75–85, and rule 6: 90–100. For example, if an
agent at a particular time has a repertoire of imaginable behavior that
consists of rules 2, 4, and 5, it means that the number representing its
behavior can only come from the ranges 30–40, 60–70, or 75–85.
Agents select their behavior from only one of their three rules. Because
they seek to adapt to their environment, this is the rule that has proven
to be closest to the collective behavior. Agents keep the other two rules
private – that is, they do not reveal them to the remaining agents in
their society.
The model’s logic of social appropriateness posits that agents choose
their behavior in line with the behavior they observe in the population.
Since agent behavior in the model is driven by rules, agents choose
their rules based on how well they match observed collective outcomes
in the agent population. That is, rules that have proven successful in
keeping the agent close to the behavior of the population are retained,
whereas others are suppressed. The model incorporates this by letting
agents keep “success scores” for each of the three rules they currently
have in their repertoire. Agents reward or penalize each rule – both the
public and the private ones – according to how close the rules turn out
to be to the collective outcome, which is simply the average behavior
across all agents (the mean across all chosen behaviors, each in the
186
Martin Austvoll Nome & Nils B. Weidmann
[0,100] interval). From time step to time step, agents behave according
to the rule with the best score, and at set intervals they replace the
worst performing rule with a fresh rule from the universe of seven
rules.3 In this way, social identities change or are habituated in a
process where norms bias choice. That is, certain behaviors are discounted as inappropriate for one’s identity (Abdelal et al. 2009, 21).
In Hoffmann’s model, however, perception of the collective outcome
is not perfect. That is, agents do not use the true average behavior of
the population, but a slightly blurred value of it. This implementation
is meant to capture the uncertainty that surrounds agents’ perceptions.
Hoffmann presents two arguments for why this imperfect social perception is likely to occur: First, lack of information, since some agents
may be wrongly perceived or not at all, and second, the complexity of
the social environment which makes it difficult for agents to assess the
predominant behavior (Hoffmann 2008, 193). When people develop
an image of the society in which they live, they do so based on clues
from selective sources, and then aggregate up. This process of aggregation leads to truer images in less complex societies and distorted images
in more complex societies. Systematic distortion such as media bias
and censorship contribute to complexity. The level of noise is implemented in the model as a confidence interval around the true population behavior: if noise is high – and the interval large – agents perceive
a randomly chosen value from within this interval, rather than the true
outcome.
When norm entrepreneurs are present, that is, when certain agents
seek to define social identities in particular ways, they enter the model
by proposing a rule that agents adopt with a certain probability, and
privately. For example, if a norm entrepreneur promotes rule 6, some
agents adopt the behavioral range 90–100 as one of their two private
rules. This has no immediate impact on collective behavior, but a norm
that is latent for some agents from that point can emerge via social
adaptation. Norm entrepreneurship thereby involves the simultaneous
altering of agents’ repertoires of imaginable behavior. Agents behave
3
The fact that we program our agents to keep “success scores” and follow rules
according to how well they perform is not intended to suggest they are
instrumentally rational. Our agents are norm following. We simply need some
way to operationalize the difference between more and less appropriate behavior.
Following Hoffmann in attaching scores to the different rules is a sensible way of
operationalizing appropriateness.
Conflict diffusion via social identities
187
according to the new rule – and thereby make it public – only if they
expect it to be the best approximation of collective behavior.
As an illustration of norm entrepreneurship, consider an example
from Russia’s Tatarstan, where the Tatar Public Center (TOTs) promoted the revival of Tatar language (Giuliano 2000). Formed in 1988,
TOTs acted as a norm entrepreneur in trying to elevate Tatar as the
appropriate language among urban, Russian-speaking Tatars. Some
Tatars equated the loss of the Tatar language with the cultural death of
their nation, suggesting that they considered language as an important
constitutive norm of their Tatar social identity (Giuliano 2000, 305).
In the end, TOTs was unsuccessful in its norm entrepreneurship.
Tatarstan experienced no wholesale revival of Tatar. Instead, urban
Tatars continued adhering to an existing and competing constitutive
norm of an alternative social identity, speaking Russian as a marker of
higher social status (Giuliano 2000, 306). TOTs’ nationalist program
failed, in part because its norm entrepreneurship only resonated with a
small portion of its target audience (305–306). It is a reminder of an
important element in the way we model norm entrepreneurship: the
reach of entrepreneurs may be limited, and we cannot take for granted
that norm entrepreneurs reach all individuals in the population and
affect their latent norms in the same way. As we model norm entrepreneurship, we therefore vary the receptivity among the population, and
examine how this affects the success of the mobilization effort.
By implementing the model of social adaptation and norm entrepreneurship as we have so far described it, one would simply replicate
Hoffmann’s (2008) model of norm emergence in a single society. Social
adaptation and norm entrepreneurship would operate in a closed
polity. The model would capture neither transnational relations nor
cross-border diffusion. To explore transnational diffusion, we add to
Hoffmann’s model by letting two societies evolve side by side. We add
a second country and transnationalize the context in which social
adaptation and norm entrepreneurship are at work.
The transnational diffusion of social identities
How can constitutive norms of social identities spread from one country to another? As described above, Hoffmann’s (2008) model simulates the emergence of norms among a set of agents in only one
country. In this section, we describe a simple extension to the model
188
Martin Austvoll Nome & Nils B. Weidmann
that introduces different mechanisms of norm diffusion across national
boundaries. In a stylized setting of two countries – each following the
Hoffmann model – we ask in what particular ways norms can spread
from one country to another. We designate one country (the source
country) to be the origin of a behavioral pattern, and a second country
(the recipient country) to be the potential follower of this pattern.
Once we transnationalize the mechanisms of social adaptation and
norm entrepreneurship, either mechanism might generate outcomes
that are observationally indistinguishable: the emergence of a norm
in the recipient country following the emergence of the same norm in
the source country. However, since we can control the mechanisms
behind this convergence, we can examine how and under what conditions convergence is reached.
In our stylized model, norms move, but people do not. Even though our
model treats the populations of the two countries as separate, it also
applies to cases where migration creates overlaps between the two populations. For example, in the case of the PKK in Germany discussed above,
migration from Turkey into Germany led to a diffusion of ideas across
borders simply by means of the individuals holding these ideas. Similarly,
Harpviken and Lischer (this volume) explain post-return violence as the
result of refugees being socialized in a violence-breeding context abroad,
and bringing home the corresponding norms and ideas. In short, migration could partially explain why the population in the recipient country
adopts norms from the source country, or why norm entrepreneurs in the
recipient country advocate these norms. We acknowledge the importance
of migration as a diffusion mechanism. However, in this chapter we take
advantage of agent-based modeling to explore the much harder-toobserve diffusion of norms when people do not cross borders.
Hoffmann’s base model allows for collective behavior along a
dimension between normal and extreme. Whereas normal behavior
at the center of the 0–100 scale is most frequent because of the
averaging of agents’ behaviors, there are occasional instances of
extreme behavior at the upper and lower ends. We set up our model
such that the source country converges to extreme behavior, and
examine how this behavior can spread to the recipient country. The
reason for focusing on extreme behavior is twofold. First, theoretically,
the aim of our paper is to explain how a non-normal behavior – strong
group antagonism – can spread between countries. Diffusion is defined
such that the origin of this process (our source country) must show this
Conflict diffusion via social identities
189
behavior before it can travel to the recipient country. Second, methodologically, the Hoffmann model’s natural tendency towards normal
outcomes at the center of the scale would not allow us to attribute the
occurrence of this behavior in both countries to the mechanisms we are
examining, since they may simply be due to independent normal
outcomes in both countries.
We generate extreme behavior in the source country by inserting a
norm entrepreneur into the population that advocates an extreme rule at
the lower end of the scale (rule 0, 0–10). This leads to a convergence to
this rule in the source country. In the following, we explore social adaptation and norm entrepreneurship, either of which might lead to the
recipient country adopting this norm, which corresponds to a diffusion
of extreme behavior across national borders. We proceed in three steps.
First, we describe a baseline setting in which the two countries have no
links with each other, which makes diffusion impossible. Second, we
transnationalize the context of social adaptation by linking the agent
populations in the two countries, which can lead to the spread of extreme
norms by mass-driven diffusion. Third, we examine the elite-driven
mechanism of norm entrepreneurship that might generate the spread of
norms. In our artificial world, we examine the impact of these diffusion
mechanisms under two conditions: a scenario of perfect social perception
(low noise level) and a scenario of fuzzy perception (high noise).
Countries without linkages
Our first baseline test illustrates the model setup for a pair of disconnected countries. We do so to provide a contrast against which to
compare the diffusion models. It also allows us to illustrate the impact
of variation in noise levels, independent of the diffusion mechanisms.
We create two countries, the source and the recipient country, and
populate them with 10 agents each. In order to generate extreme
behavior in the source country, we insert a norm entrepreneur who
advocates rule 0 (0–10). As a result, most of the model runs lead to a
norm between 0 and 10 in the source country. Since we do not specify
any linkages between the two countries, we expect the outcomes in the
two countries to be unrelated to each other. In particular, there will be
no convergence of the recipient country norm to the extreme source
country outcome. Figure 7.1 illustrates this baseline model for two
sample runs, under the low and high noise scenarios.
100
100
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
0
25
50
75
100
125
150
175
200
0
25
50
75
100
125
150
175
200
Figure 7.1 Single model runs with two disconnected countries, for low (left panel) and high noise levels (right). The lines
show norm emergence over time in the source country (grey) and the target country (black).
Conflict diffusion via social identities
191
Table 7.1 Number of runs (out of 50) with successful norm convergence
in the baseline model
Low noise
High noise
4
0
The grey line in the left panel shows that following a period of
unstable norms, the source country converges to an extreme norm, as
advocated by the norm entrepreneur. The recipient country is not
affected by developments in the source country. It quickly stabilizes
around normal behavior (around 50). In the high noise scenario (right
panel) the outcome is the same for the recipient country. However,
fluctuations are much higher for the source country, where – as shown
in Hoffmann (2008) – deviations from the norm proposed by the
entrepreneur are more drastic and occur frequently when noise is high.
The plots above show results only for single model runs. However,
in each agent-based model, a number of probabilistic moves are made.
In our model, for example, agents do not perceive the true collective
behavior, but rather a noisy estimate of it. This noise is essentially
random; we only control its magnitude using the noise parameter. In
order to make sure that these random outcomes do not determine the
outcome in the single runs, we repeat the model for different sequences
of random numbers. Since we cannot easily eyeball all these runs to
check for convergence, we need to define a criterion so that the computer can calculate whether in a particular run, the norms in the source
and recipient country have converged. We therefore define norm convergence to have occurred once the difference between the true collective behaviors in both countries is 15 or less for 200 consecutive time
steps. This threshold is arbitrary, but changing it within certain limits
does not alter our conclusions. Each model run ends when convergence
is reached, or otherwise, is terminated at time 3,000. Table 7.1 reports
the number of model runs (out of 50) where we observe convergence.
The table shows that when both countries are disconnected, there is
very little if no convergence at all. For the low noise scenario, there are
a few runs with seemingly successful diffusion, but these are due to the
fact that neither the source nor the recipient country is hard-wired to
extreme or normal behavior. For example, rather than settling on a
192
Martin Austvoll Nome & Nils B. Weidmann
norm around the center, the population in the recipient country – while
completely disconnected from source country – can settle on an outcome close to 0, which our model counts as convergence. However,
these instances are rare and do not challenge our general approach.
Transnationalized social adaptation
We are now ready to apply our first diffusion mechanism to the baseline
model. Given our theoretical considerations, we argue that one route by
which norms spread is via transnational intra-group networks where, in
effect, the social context of social adaptation is transnationalized. We
translate this to our model by letting recipient country agents base their
assessment of their society’s norm in part by observing the population in
the source country. In other words, the individuals in the recipient country rely on information from the source country to determine which rule
in their repertoire of imaginable behavior is the most appropriate.
The social contexts to which agents adapt are transnationalized once
networks and means of communication cross national boundaries. For
example, Haider (2005) describes how the transnational context of
Xinjiang’s Uighur population was transnationalized when the
Karakoram Highway opened in 1982, linking Kashgar, Xinjiang with
Islamabad, Pakistan. Uighurs are predominantly Muslim, and therefore
share faith with the majority Pakistani population. Many Uighurs traveled to Pakistan to study in religious schools well into the 1990s. Their
time in Pakistan contributed to instilling in them a “strong, moderate
Muslim identity” (Haider 2005, 529). As the Islamic awareness grew
among the Uighurs, one might expect that this process involved the
adoption of the constitutive norms of Muslim identity. Much like the
emergence of Kurdish identity among Turkish immigrants to Germany
depended on the relative liberties of German society, the emergence of a
stronger Muslim identity among Uighurs occurred at a period of
liberalization in China. Transnational relations with Muslims in
Pakistan were facilitated by the greater economic freedoms, the encouragement to trade with neighbors, and the greater cultural and religious
liberties in China of the 1980s (Haider 2005, 525).
This underlines another assumption in our model: there is no strong
state or other third party that by means of coercion can control agents’
actions. The Karakoram Highway was to some extent conducive to
conflict between Uighurs and China. Some Uighurs who traveled to
Conflict diffusion via social identities
193
Pakistan took part in the violent resistance against the Soviets and in
turn the US in Afghanistan (Haider 2005, 529). Upon their return to
Xinjiang, they took part in the violent resistance against China as part
of the nationalist movement (Haider 2005, 530). Recent unrest in
Xinjiang testifies to the continuation of social conflict. Xinjiang’s social
cleavages continue to manifest themselves in the constitutive norms of
social identities – some very particular. As Haider describes one norm,
“feelings of alienation remain strong, with some Uighurs shunning
Han restaurants because they serve pork, while looking down on those
Uighurs who do eat at Han establishments” (Haider 2005, 527).
The opening of the Karakoram Highway was one manifestation of
the transnationalization of a social context. In our model, this incorporation of transnational information is guided by a single parameter,
transnational impact, which controls the extent to which the source
country collective behavior affects agents in the recipient country. This
impact parameter takes values between 0 and 1. For example, if transnational impact is set to 0.4, the recipient country agents’ perception of
the appropriate collective behavior will be the weighted average of the
source country behavior (40%) and the behavior in their own country
(60%). The baseline model as described above is a special case of this
one with transnational impact set to 0. Figure 7.2 again shows the norm
trajectories in both countries, for two selected runs of the model.
The left panel in Figure 7.2 shows the quick effect of transnationalized social adaptation under low noise conditions. Already after a
small number of time steps (about 50), the dominant norm in the
recipient country converges to the extreme outcome in the source
country. Yet, there is no perfect overlap. Because of the model’s natural
trend towards the center, norms in the recipient country turn out to be
somewhat closer to the center than the extreme outcomes in the source
country. The disparity between the points at which source country and
recipient country converge suggests that local conditions are important
in determining the outcome of diffusion. Whereas the extreme outcome
in the source country is driven by a norm entrepreneur, this entrepreneur promotes a rule that is not adopted without alteration by the
recipient country population. Instead, the emergent norm in the recipient country has its own local flavor. The imperfect convergence echoes
Schmitz’ (this volume) argument that local factors shape the impact
of transnational mechanisms. Schmitz analyzes how one insurgent
organization adopted a transnational (Christian) ideology – “a global
100
100
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
0
25
50
75
100
125
150
175
200
0
50 100 150 200 250 300 350 400 450 500 550
Figure 7.2 Diffusion by transnationalized social adaptation, for transnational impact set to 0.4, under the low and high
noise scenario.
Conflict diffusion via social identities
195
structure of values” – and put it to work to address local grievances
stemming from economic and political marginalization.
An analogy in IR work on transnationalism is the phenomenon of
norm localization (Acharya 2004). Acharya defines localization as
“the active construction . . . of foreign ideas by local actors, which
results in the former developing significant congruence with local
beliefs and practices” (Acharya 2004, 245). One should be careful
not to belabor this analogy. We do not explicitly model norm localization. Instead, our recipient agent society settles around a norm that is
influenced by a very generalized natural attractor. Still, Acharya’s
work on norm localization and the literature that has emerged in its
wake (see, for example, Williams 2007 and Capie 2008) point to some
ideas that could well be incorporated in subsequent agent-based
models. Such ideas include dynamics of contestation between outside
and inside entrepreneurs, the distinction between inside and outside
initiatives for diffusion, and the connection between social identities
and institutions in the recipient country. We return to some of these
ideas in the conclusion.
In Figure 7.2, the picture changes once we introduce noise. The
recipient country norm follows the lead given by the source country,
but because of the fluctuations due to noise, convergence does not
occur. This result suggests that diffusion by social adaptation can only
occur once a stable norm has emerged in the source country that agents
in the recipient country can follow. If this is not the case, reliance on a
volatile example set by the source country can even disturb the modest
norm emergence in the recipient country that would have occurred had
there been no transnational linkages.
Again, we examine the success rate of norm diffusion on a larger set
of cases. We run the model for different levels of transnational impact
(0.1, 0.2, 0.4) and count the number of successful diffusions under low
and high noise. We intentionally keep the level of impact low (at least
below 0.5), because we believe that the domestic dynamics carry much
more weight in the norm emergence process than outside influences.
Table 7.2 reports the results. Two findings stand out. First, diffusion
by social adaptation can only be successful under low noise conditions.
Very few runs in the high noise scenario exhibit norm diffusion.
Second, there is a non-linear effect of the transnational impact parameter. Low values of transnational impact are insufficient to generate
diffusion. However, as we increase this parameter to 0.4, all of our
196
Martin Austvoll Nome & Nils B. Weidmann
Table 7.2 Number of runs (out of 50) with successful norm diffusion
for different levels of transnational impact
Transnational impact
Low noise
High noise
0.1
0.2
0.4
0
3
50
0
1
3
model runs exhibit successful diffusion. Clearly, small shifts in the
transnational impact parameter can dramatically change the effect of
social adaptation.
Transnationalized norm entrepreneurship
Next, we consider norm entrepreneurship as a mechanism of diffusion.
As we argue in the theory section above, entrepreneurs can be key
actors for the collective establishment of norms, as they actively propagate a certain norm. Recall that in Hoffmann’s model, an entrepreneur’s suggestion only affects the agents’ set of private rules. There is
no guarantee that this suggestion will eventually become the rule by
which all agents behave. Hence, the norm emergence that the model
generates is the result of a constant shaping of the agents’ repertoires of
imaginable behavior and subsequent social adaptation. In the
following, we examine how this shaping can account for the diffusion
of norms between the two countries.
In the previous section, we modeled diffusion purely by transnationalizing the context in which agents adapt, with no norm entrepreneur
present in the recipient country. We now implement a simple
entrepreneur-driven mechanism of diffusion by introducing a norm
entrepreneur into the recipient country. This entrepreneur advocates
the predominant norm from the source country whenever agents
update their norm repertoire (on average at every 25th time step). It
is important to note that as mentioned above, agents only add this rule
to their existing repertoire, from where it may be activated only if
evaluated favorably. In addition, norm entrepreneurs might be limited
in their success in changing agents’ repertoires, which is what we
observed in the Tatarstan example. We therefore introduce a
receptivity parameter into our model, which captures the probability
Conflict diffusion via social identities
197
Table 7.3 Number of runs (out of 50) with successful entrepreneurdriven norm diffusion
Receptivity
Low noise
High noise
1
0.9
0.8
0.7
0.6
0.5
21
28
20
16
12
12
20
20
16
14
8
3
that an agent incorporates the norm promoted by an entrepreneur into
his or her own repertoire. For example, with receptivity set to 0.8, on
average only 80 percent of the population would include the proposed
rule, whereas the remaining 20 percent would simply ignore it.4
This extremely simple implementation shifts the locus of diffusion
from the population as in the previous experiment to the norm entrepreneurs in the recipient country. Do we observe norm diffusion, even
though the agents themselves have no direct perception of the happenings in the source country? Figure 7.3 shows the results of two model
runs using our entrepreneur-driven mechanism, under the low and
high noise scenarios and with receptivity set at 100 percent.
With little noise present, the recipient country converges to an
extreme norm quickly and remains there, even though the source
country is undergoing some initial fluctuations before settling on the
same outcome. In essence, norm entrepreneurship seems to work,
similarly to social adaptation as shown above. What if we now impose
high noise on agents’ perceptions? Whereas high noise largely eliminates diffusion by social adaptation, this does not seem to be the case for
norm entrepreneurship. High noise introduces larger fluctuations
during the beginning of the simulation, but does not impede the recipient country’s convergence to the extreme behavioral norm 0–10. Still,
we need to examine if these findings are robust even if entrepreneurs
are only partially successful, which is why we examine a larger set of
runs that vary both receptivity and noise level. Table 7.3 reports the
4
Since in the model we are interested only in the recipient country dynamics,
the success rate is only applied to the entrepreneur in that country.
100
100
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
0
50
100
150
200
250
300
350
400
0
50
100
150
200
250
300
Figure 7.3 Diffusion by transnationalized norm entrepreneurship, under the low and high noise scenario.
350
Conflict diffusion via social identities
199
number of successful diffusions in a set of 50 runs with different
random number sequences.
The results in Table 7.3 support our initial conclusions. In contrast
to our above findings for transnationalized social adaptation,
entrepreneur-introduced norms from the source country can successfully be established in a significant number of runs, regardless of
whether we apply low or high noise. These results hold even if we
assume that entrepreneurs are only moderately successful (receptivity
at 0.6–0.7). In other words, whereas perfect social perception (low
noise) is a precondition for social adaptation, entrepreneur-driven
diffusion is much more robust to varying levels of noise and works
almost equally well under both conditions, even if entrepreneurs have
only limited reach.
Discussion and conclusion
The importance of norm entrepreneurship is supported by at least two
empirical contributions to this volume: Bakke’s (this volume) study of
goals, tactics, and resources in the Chechen wars, and Adamson’s (this
volume) work on diaspora mobilization in the PKK – Turkey conflict.
Bakke describes training camps in Chechnya as sites of the transnational diffusion of goals and tactics. A central piece of evidence
involves one transnational agent of diffusion, Khattab, setting up a
training center in Chechnya in which young recruits were trained in
religion and military tactics by foreign teachers (Bakke, this volume).
These transnational insurgents sought to instill their beliefs in the
young recruits, or to convey ideas from foreign ideologues, and in this
we see norm entrepreneurship at work.
Adamson analyzes how the social identity of Turkish immigrants in
Germany changed into a Kurdish insurgent consciousness. She argues
that PKK activists worked to promote a particular Kurdish identity in a
way that closely resembles our understanding of norm entrepreneurship. In our agent-based model, norm entrepreneurs affect people by
making available new norms that could influence subsequent behavior.
Similarly, when PKK activists promoted particular constitutive norms
such as language and customs, they “carved out a publicly available
alternative identity category” to either “German” or “Turkish”
(Adamson, this volume, emphasis added). In other words, political
entrepreneurs formulated “political categories and ideologies that
200
Martin Austvoll Nome & Nils B. Weidmann
presented individuals with identity options that countered both
German and Turkish national identities” (Adamson, this volume,
emphasis added). Adamson thus describes the diffusion of a separatist
social identity from Turkey to Germany through a process where norm
entrepreneurship was important.
At the same time as Bakke and Adamson’s contributions lend empirical plausibility to the mechanism of norm entrepreneurship, our agentbased model suggests extensions to their work that could be explored in
future research. Recall that we model norm entrepreneurship not as a
stand-alone mechanism, but as a sequence of transnational norm entrepreneurship and domestic social adaption. The norm entrepreneur in
the recipient country is a transnational agent because he or she promotes
the current norm from the source country. However, for that norm to
translate into behavior in the recipient country, it depends on a process
of social adaptation within the boundaries of the recipient country.
As a possible contribution to Bakke’s work, our computer simulations suggest social adaptation as a mechanism that could account for
the dynamics of training centers such as that established by Khattab in
Chechnya. A presumably transparent social setting of young recruits
would be conducive to social adaptation. Norm entrepreneurship
would work in sequence with social adaptation where transnational
insurgents sought to instill their beliefs in the young recruits. Indeed,
our simulation suggests that the product of such training centers need
not only be a change in the framing of an insurgent movement, as
Bakke argues, but a more fundamental change in its social identity.
Our model of transnational norm entrepreneurship could contribute
to Adamson’s account of identity change by adding social adaptation
as an account of how these alternative identities might emerge intersubjectively in a population. As a stand-alone mechanism, norm entrepreneurship is an insufficient account of identity change. It misses
important social dynamics that are involved when new identity norms
arise. Instead, a sequence of norm entrepreneurship and social adaptation capture a more complete process of identity change.
To summarize, our agent-based model suggests that both social
adaptation and norm entrepreneurship are candidate explanations of
diffusion. When the social context to which agents adapt is transnationalized, norms that constitute social identities can diffuse from one
country to another. Alternatively, norm entrepreneurs who work as
transnational agents can promote norms from a source country to a
Conflict diffusion via social identities
201
recipient country audience, leading to diffusion. Either mechanism can
operate in societies that are both more or less complex. Herein lies the
difference. Norm entrepreneurship leads regularly to diffusion irrespective of the level of social complexity. Social adaptation leads to
diffusion only in more transparent societies. In short, norm entrepreneurship is the more robust mechanism of diffusion.
As the process of diffusion is currently operationalized, it is meant to
be a simple base model for norm diffusion in a transnational context and
leaves ample room for extensions. For example, an interesting question
to explore is the interaction of social adaptation and norm entrepreneurship. What if the two mechanisms complement each other, and what if
they compete? Similarly, another form of competing diffusion mechanisms could occur between norm entrepreneurs in different locales. Work
on norm localization points to processes where outside and inside norm
entrepreneurs contest each other (Acharya 2004). Implementing this
process would require us to introduce entrepreneurs that target the
recipient country population both from within and from a source country. Finally, our model could be extended to include abstract representations of political institutions in the recipient country, in an effort to
understand how norms shape these institutions, but also how institutions affect the emergence and diffusion of behavioral norms.
part iii
Theory, mechanisms, and the
study of civil war
|
8
Causal mechanisms and typological
theories in the study of civil conflict*
andrew bennett
Introduction
Most political scientists are rightly focused on their empirical and
theoretical research and its policy implications, but this has often left
them out of date in their understanding of debates in the philosophy of
science. Many political scientists have continued to espouse ideas from
scholars like Carl Hempel, Thomas Kuhn, and Imre Lakatos that have
long been considered deeply problematic among philosophers of science. A recent survey of American scholars in the international relations subfield, for example, found that approximately two-thirds
identified themselves as “positivists” (Maliniak et al. 2011, 454), even
though the philosophical schools of thought most closely associated
with this label fell out of favor among philosophers of science decades
ago. Other notions still common in political science but long since
discarded by philosophers include explanation by reference to
“covering laws” (Waltz 1979), self-styled Kuhnian “paradigm” wars
in the international relations subfield (Maliniak et al. 2011), and
efforts to construct theoretical schools of thought as Lakatosian
“research programs” (for critiques see Elman and Elman 2003; Sil
and Katzenstein 2010b).
The epistemological challenge for political scientists is that the remnants of these earlier philosophies of science are at odds with an
increasing focus on explaining complex phenomena via reference to
“causal mechanisms.” The idea of explanation via reference to causal
mechanisms builds on the work of “critical realists” like Roy Bhaskar
(1975) and more contemporary “scientific realists” like Colin Wight
(2006, 2007; on the differences between scientific and critical realism,
see Chernoff 2007). One need not take recent versions of scientific
realism as unproblematic – indeed, they are not (Chernoff, 2002) – to
* I thank the contributors to this volume, and particularly its editor Jeff Checkel,
for insightful comments on several drafts of this chapter.
205
206
Andrew Bennett
argue that scientific realism offers important improvements over earlier
philosophies of science and that it has pragmatic value for a discipline
focused on explaining, understanding, and to the extent possible
predicting and influencing the complexities of political life. This chapter explores the costs and benefits of explanations that reference causal
mechanisms, pointing out how the mechanism approach to explanation differs from earlier philosophies of science. It draws upon the
empirical chapters in this volume, as well as on the broader literature
on civil war, to illustrate the benefits of focusing on explanations that
invoke hypothesized causal mechanisms.
The chapter proceeds as follows. First, I define causal mechanisms
and distinguish explanation via reference to causal mechanisms from
both covering law explanations and post-modernist approaches to
understanding. Second, the chapter addresses the relationship between
the epistemology of explanation via mechanisms and common
methods of political science research, focusing on the method of process tracing but including a brief discussion of statistical methods and
formal modeling. The chapter then presents a taxonomy of some of the
most commonly theorized mechanisms in political science including,
but not limited to, the mechanisms invoked in this volume’s empirical
studies. I illustrate the value of this taxonomy by using it to classify the
mechanisms theorized in the empirical chapters within this volume and
to identify the important categories of mechanisms that these chapters
leave out. This leads – finally – to a discussion of typological theorizing
as a means of addressing combinations of mechanisms and the complexities this creates, with a focus on how to build such theories within
the literature on civil war. I conclude that explanation via reference to
causal mechanisms entails important costs, particularly a substantial
loss of parsimony relative to most extant political science theories, but
I argue this is the necessary price for improving our understanding of
the complexities of political life.
Causal mechanisms – definitions and theoretical framing
In his introduction to this volume, Checkel borrows from John Gerring
to define causal mechanisms as “the pathway or process by which
an effect is produced or a purpose is accomplished” (Checkel, this
volume, citing Gerring 2007b, 178). This is a good start, especially
with Checkel’s elaborations that causal mechanisms are “ultimately
Causal mechanisms & typological theories
207
unobservable ontological entities that exist in the world, not in our
heads.” This distinction, easily overlooked, is critical, and deserves
contextualization within the history of the philosophy of science. The
appeal to metaphysical unobservables is precisely what the positivists
wanted to avoid, but their efforts to elide this appeal created sharp
contradictions (Salmon 1990). The positivists’ conundrum can be
summed up in the bumper sticker slogan that “observation is theoryladen.” We cannot even take in the direct stimuli that our senses
receive without filtering them through implicit or explicit theories that
try to make sense of these stimuli.
Recognition of this problem led to a search for justifications of
theories rather than observations as the bedrock of explanation. This
motivated the attempt by Carl Hempel and Paul Oppenheim to justify
theories by reference to “laws” or “covering laws,” known as the
“deductive-nomological” approach to explanation. In this view, to
explain an outcome was to refer to a covering law, or an underlying
regularity, of which the outcome was an instance. This effort at explanation foundered on the inability to find any independent warrant for
covering laws – something Hempel and Oppenheim promised in a
famous footnote (1948 [Hempel 1965, 273]), but ultimately failed to
deliver.
Subsequently, the “scientific realist” school substituted the
assumption of an ontological reality independent of our minds and
our theories for Hempel’s search for justification via “laws.” Our
theories, in this view, aspire to refer to this independent reality with
as much verisimilitude as possible. This may not seem a very heroic
assumption, but it presents six philosophical challenges that merit
discussion in view of this volume’s emphasis on explanation via reference to causal mechanisms.
First, by what standards can we judge whether one theory about
causal mechanisms has greater verisimilitude than another? Imre
Lakatos (1970) focused on this challenge of judging scientific progress.
Lakatos dismissed as “naïve falsificationism” the idea that theories
could be readily falsified by inconsistencies with observed “facts,” as
measurement errors and other problems could also account for failed
predictions. Lakatos objected even more strongly to Kuhn’s (1962)
suggestion that the most relevant standard for progress was the consensus of the relevant scientific community, a view that Lakatos found
too subjective. Instead, Lakatos argued that theories could be judged as
208
Andrew Bennett
progressive if over time they uncovered “novel facts” that were empirically verified. Later discussions (Elman and Elman 2003) clarified that
two kinds of novelty are key: use novelty, or the ability of theories to
predict facts that are independent of the evidence from which these
theories were derived, and background theory novelty, or the ability of
theories to predict facts that are inconsistent with alternative theories.
Although other aspects of Lakatos’s approach have undergone sharp
criticism, these two kinds of novelty remain as useful albeit fallible
standards for assessing whether our theories about how mechanisms
work are progressive (Bennett 2003).
Second, what are we to make of the metaphysical appeal to
ontological entities that are ultimately unobservable? In one sense, this
is no more problematic than the positivists’ reference to unobservable
“laws.” After all, most of what we “know” about the political and
social world we learn not through direct observation but through
reading, hearing, or seeing books, journals, and news reports. Some
of these sources and their instruments of observation are more reliable
and less susceptible to user biases than others, and we rightly do not
treat all observations as being created equal.
It is useful here to think of a movable border or horizon between the
observable and the unobservable worlds (George and Bennett 2005).
As our instruments of observation improve, as in the introduction of
public opinion polls or brain scans, we push back the border of the
unobservable world. Yet ultimately there always remain more discrete
or distant mechanisms that we cannot directly observe. Still, as scientific realists maintain, our theories about mechanisms often generate
observable implications on what should be true if the posited mechanisms operate in the manner that we theorize. We can test these implications to assess the accuracy of our theories, even if we cannot
observe mechanisms or causation directly or unproblematically assess
how well our theories fit the observable evidence.
Third, how does explanation via reference to mechanisms differ
from explanation via reference to “laws”? In some respects, the two
seem quite similar: to explain an outcome in either sense is to show that
it was to be expected under the circumstances. There is a crucial
difference, however, in that deductive-nomological or covering law
explanations admit as “causal” many relationships that a mechanism
view excludes. The covering law model, for example, allows the readings on a barometer, which provide a strong correlation with the
Causal mechanisms & typological theories
209
weather, to be accepted as an explanation of the weather. The problem
is that this form of explanation readily adopts “as if” assumptions: in
this view, it does not matter if entities acted as the theory suggests, it
only matters that outcomes behave “as if” this were true (Friedman
1953). A mechanism-based approach to explanation, on the other
hand, eschews “as if” assumptions and requires, in principle, that
our theory about how underlying mechanisms work must be consistent
with the finest degree of detail that we can observe.
Fourth, and related, does a commitment to explanation via causal
mechanisms require all social scientists to become methodological
individualists, or even neuroscientists? On the one hand, requiring that
our theories must be consistent with what we can observe at lower
levels of analysis does not mean that the interesting explanatory action,
the most cost-effective means of theory testing, or the relevant policy
levers are at these lower levels of analysis. If a social structure is so
strongly confining and self-reproducing that all individuals in it behave
the same, then the interesting explanatory leverage is at the level of the
social structure, not that of an individual’s brain chemistry. On the
other hand, a commitment to explanation via mechanisms does make
theories suspect if it can be shown that individual actors did not make
the calculations or engage in the behaviors posited by structural
theories. Ultimately, assumptions that are more accurate on the micro
level should lead to more accurate structural or systemic models.
Fifth, is explanation via reference to mechanisms incompatible with
constructivist or post-modern views of social and political life? The
short answer is that the lines of thinking espoused by some constructivists and post-modernists are irreconcilable with explanation via
mechanisms. At the same time, those constructivists who have
embraced scientific realism (Wendt 1992) are comfortable with
mechanism-based explanations. As Checkel makes clear in his introductory chapter, scientific realism is open to all kinds of mechanisms
that constructivists emphasize, including persuasion, learning, naming
and shaming, framing, legitimacy, and appropriateness, as well as
mechanisms involving material power and transactions costs.
Scholars who are skeptical of any possibility of causal explanation,
however, will remain skeptical of mechanisms. This includes researchers who believe that social agents and social structures are so inextricably mutually constituted “all the way down” to the finest observable
slices of space and time that there is no sensible way of separating the
210
Andrew Bennett
two into causes and effects. Many constructivists and post-modernists
are not unalterably opposed to aspirations to causal explanation,
however, and leading interpretivists suggest in their writings that
behavior can be explained and that some explanations are better than
others (Hopf 2007). Scientific realists remind us that although observation is theory-laden, it is not theory-determined. The social world
that exists independently of our minds can surprise us and force us to
rethink our theories and explanations.
Sixth and finally, what other accounts of causation and explanation
are consistent with or incompatible with explanation via reference to
causal mechanisms? Henry Brady has very usefully summarized three
approaches to causal explanation, which he terms neo-Humean regularity theory (focusing on constant conjunctions), counterfactual
theory, and manipulation theory; these are in addition to the
mechanism-based understanding presented here (Brady 2008).
My argument is that the mechanisms approach is compatible with
all of these except the positivist variants of neo-Humean regularity
theory. The counterfactual approach, which Brady associates with
statistical methods and with efforts to establish and measure “causal
effects,” is fully compatible with a search for theories about the mechanisms that might account for such effects. It makes no sense to discuss
effects that lack mechanisms or mechanisms that lack effects (see also
Checkel’s introduction to this volume). Similarly, a “manipulation”
account of explanation and inference that emphasizes the role of
experiments and the ability to manipulate potential causes is perfectly
compatible with explanation via mechanisms. Experimentalists typically have in mind or would like to uncover theorized mechanisms that
might account for the outcomes that result from the manipulation of
specific variables.
These six issues are what led Alex George and I to adopt a definition
of causal mechanisms that is more detailed than Gerring’s but not
incompatible with it. We defined causal mechanisms as “ultimately
unobservable physical, social, or psychological processes through
which agents with causal capacities operate, but only in specific contexts or conditions, to transfer energy, information, or matter to other
entities,” thereby changing the latter entities’ “characteristics, capacities, or propensities in ways that persist until subsequent causal mechanisms act upon it” (George and Bennett 2005, 137). This definition
places mechanisms on the ontological level. It also makes clear that
Causal mechanisms & typological theories
211
theories about mechanisms may not be universal laws that operate in
all times and places, but may instead be operative in some circumstances and not others, and that their operation is mediated by that of
other mechanisms.
Mechanisms and methods
All of the three most common general methods in political science –
statistical analysis, formal modeling, and case studies – can contribute
to the development and testing of theories about causal mechanisms.
Statistical methods can test whether any population-level observable
implications of hypothesized mechanisms are borne out. Formal
models can deductively drive a researcher to theoretical insights about,
or testable implications of, hypothesized mechanisms that the
researcher otherwise might have missed (see also the discussion in
Nome and Weidmann, this volume). This chapter focuses mostly on
case study methods for identifying and testing theories on causal
mechanisms, particularly process tracing. Process tracing in individual
cases provides a powerful means of using both induction and deduction to develop and test theories about hypothesized mechanisms. Case
studies afford an opportunity to examine closely sequences of actions
and events, and the details that emerge, often unknown even to experts
on the case prior to its intensive study, provide many opportunities to
test the observable implications of alternative explanations. The
detailed analysis of sequences also provides some leverage over the
issue of causal direction – that is, did A cause B or did B cause A?
As Checkel argues in his introduction, process tracing is the analysis
within a single case of the sequence of events that intervenes between
the starting point and outcome of the case as these are defined by the
researcher. The researcher’s prior choice of theories she or he deems to
be relevant to the case drives much of the analysis, as it determines
which intervening events are of interest, when to start and end the time
period relevant to the case, which case(s) to study, what the “case” is,
and what the phenomenon or population is of which it is a case. In
addition to this largely deductive construction of the case and testing of
the observable implications of alternative explanations within it, there
can also be an inductive side to process tracing. This involves “soaking
and poking” within a case and identifying new explanations of its
outcome. These explanations may be based on theories that are fairly
212
Andrew Bennett
new or on theories that other scholars had already proposed but that
the researcher had not thought to apply to the case prior to studying it.
Contrary to the common injunction that one cannot develop a
theory from a case and test it against the same case, it is possible to
develop a theory from a case and test it against new observable
implications within that case that are different from and independent
of the evidence that helped the researcher derive the theory. Such new
implications can provide a check on confirmation bias and satisfy the
criterion of “use novelty” discussed above if the researcher had not
thought to look for them prior to developing the theory that predicts
they should be found within the case. Detectives and medical diagnosticians as well as political scientists frequently engage in this kind of
inference.
Several researchers have begun to develop explicit community standards for what constitutes good process tracing (see Checkel, this
volume; Bennett and Elman 2006; Checkel 2006; Van Evera 1997;
and Collier 2011). One criterion is that researchers should consider a
wide range of alternative explanations, treat them fairly, and do sufficient process tracing to assess their roles. A second is that one should
relentlessly pursue the most important observable implications of alternative explanations, or in Bayesian terms, the pieces of evidence that
are most likely to update the prior probabilities assigned to the truth
value of alternative explanations. A third criterion is that good process
tracing anticipates and accounts for potential biases and gaps in the
sources of evidence. Finally, to the extent possible, good process
tracing explains each of the important potential turning points in the
sequence of events leading from hypothesized causes to observed outcomes in the case.
From analytic eclecticism to structured pluralism: causal
mechanisms and the cumulation of knowledge
What are the implications for the study of civil conflicts, and the study
of politics more generally, that emerge from a shift toward mechanismbased explanations? As Charles King has pointed out, studies of ethnic
and civil conflict have already begun to move toward a more microand mechanism-based approach (King 2004). Indeed, the latter is one
of the three moves that Checkel advances to justify this book’s value
added, with the other two being an effort to draw on theories of
Causal mechanisms & typological theories
213
transnationalism and an emphasis on measuring mechanisms in action
(Checkel, this volume).
Yet there is a tension among these moves, as explanation via causal
mechanisms, with its openness to many kinds of theory and its
emphasis on interactions among different kinds of mechanisms, is at
odds with the current division of the international relations subfield
into paradigmatic “isms” (most prominently neorealism, neoliberal
institutionalism, and constructivism) and the treatment of these isms
as if they were internally complete and mutually exclusive explanations
of phenomena. As Peter Katzenstein and Nobuo Okawara have argued
(2002), the reification of the IR subfield into competing isms has
obscured as much as it has revealed, and most scholars, when pressed
to explain specific cases or phenomena, draw on more than one
paradigmatic ism. The paradigmatic isms, in their view, are best
thought of as rubrics helping scholars organize collections of hypothesized mechanisms that can be combined to construct explanations,
rather than self-sufficient and complete grand theories of politics.
Katzenstein and Okawara, and later Sil and Katzenstein (2010b), have
thus argued for an “analytically eclectic” approach that brings
together theories on mechanisms from different isms.
This is a clear improvement over formulating schools of thought in
IR theory as if they were Kuhnian paradigms or Lakatosian research
programs. Yet the “eclectic” label can easily be misinterpreted, as it
can mean either borrowing from the best of various approaches, as
Katzenstein and his co-authors clearly intend, or patchwork. This
ambiguity is not merely linguistic, as a critical question regarding
explanation via causal mechanisms is whether it provides a framework
that allows scholars to research, explicate, and teach theories about
politics in ways that are cumulative and progressive, rather than
merely creating a collection of disparate explanations of different
phenomena.
I argue that it is in fact possible to construct a taxonomy of theories
on causal mechanisms that provides a framework for cumulative theorizing and a useful checklist to ensure we are not leaving out important mechanisms from our explanations. My taxonomy draws upon
James Mahoney’s (2000) typology of sociological explanations of
institutions, which includes those rooted in power, efficiency, and
legitimacy. This tripartite division usefully mirrors the three leading
“isms” in the IR subfield – (neo)realism (power), (neo)liberalism
214
Andrew Bennett
(institutional efficiency), and constructivism (legitimacy) – but it does
not reify these approaches into Kuhnian paradigms or Lakatosian
research programs.
It is useful to think of the power and institutional mechanisms as
involving the “logic of consequences,” and the legitimacy mechanisms
as embodying the “logic of appropriateness” (March and Olsen 1984;
Ruggie 1998). The taxonomy also embodies different approaches to
solving collective action problems and providing public goods. The
“power” solution is that one or a few of the most powerful actors
provide public goods largely or wholly by themselves. The incentive for
these actors to do so is that they are sufficiently well endowed to
substantially affect the amount of public goods provided and big
enough to reap benefits from public goods that outweigh the costs of
providing them (Olson and Zeckhauser 1966). The “legitimacy” solution involves creating shared identities that make free-riding in the
provision of public goods socially inappropriate. The “institutional
efficiency” solution involves creating institutions that increase
transparency and lower transaction costs to make it easier for actors
to provide public goods, share burdens in doing so, and identify and
punish free-riders.
A second dimension of the taxonomy draws from constructivism
(Wendt 1992) and structuration theory (Giddens 1984). It captures the
four possible combinations of mechanisms through which agents and
structures interact: agent to agent, structure to agent, agent to structure, and structure to structure.
These two dimensions form the core of the taxonomy in Table 8.1.
For illustrative purposes, I have added levels of analysis, fields of study
associated with particular agent-centered and structural mechanisms,
and examples from the literature on civil war. The examples of
hypothesized mechanisms in each box are meant to be illustrative
rather than exhaustive, though it will become evident below that they
include many of the mechanisms invoked in the empirical chapters in
the present volume.
This taxonomy serves six purposes in fostering cumulative theorizing and knowledge about politics. First, it provides a checklist that
scholars can use to make sure that they are not leaving out important
alternative explanations of a phenomenon. As Checkel noted in his
introduction to this volume, in his own work on evolving European
citizenship norms, he omitted structural factors until critics brought
Table 8.1 A taxonomy of theories on social mechanisms
Agent to agent
Structure to agent
Agent to structure
Culture as enabler and
constraint
Norm entrepreneurs, Unintended evolution of social systems
framing
Resources as enabler and
constraint
Revolution
Power transitions
Functional efficiency: Emulation diffusions Evolutionary selection
Neoliberalism, logic
of consequences
Functional
competition,
innovation
Moral hazard, adverse selection
Levels of analysis
Fields of study
Individual, social
Economics
Sociology
Social psychology
System to system
Economics
Sociology
Legitimacy:
Emulation
Constructivism, logic socialization
of appropriateness
Material power:
(Neo) realism, logic
of consequences
Hegemonic
socialization
Individual, social
Systemic
Cognitive and social Economics
psychology
Sociology
Social psychology
Examples in civil war Principal agent
literature
theory (Salehyan
2010)
Collective action (Lichbach Framing (Gagnon
1998)
1994, Kalyvas
Opportunity structures
2006)
(Fearon and Laitin 2003)
Lootable resources (Collier
and Hoeffler 1998, 2004)
Structure to structure
Demographic change as an input in civil
and ethnic conflicts (Goldstone 2002)
216
Andrew Bennett
this gap to his attention (Checkel, this volume). Using the taxonomy in
its capacity as a checklist could have alerted him to the possible role of
such factors – and before his critics did so.
Second, scholars can drill down deeper into any one of the boxes in
the taxonomy, refining or adding new theories on the mechanisms in
that category, or disaggregating theorized mechanisms into different
sub-types. Much of Checkel’s framing of this book’s contribution and
Wood’s assessment in Chapter 9 focus on this kind of improvement,
seeking more detailed explanation of the mechanisms behind the
diffusion of practices in civil war. Third, researchers can develop
increasingly comprehensive historical explanations of particular cases
drawing on theories from any or all of the categories in the taxonomy.
Fourth, using statistical methods, we can develop estimates of the
magnitude of the causal effects of the variables specified by theories
about how causal mechanisms work.
Fifth, researchers can refine the scope conditions of the theories in
any of the categories, clarifying where and when or under what conditions they are strongest and weakest and specifying more clearly the
populations to which they apply. The specified populations should
ordinarily include “negative cases” that could have had the outcome
of interest but did not (Goertz and Mahoney 2004).
Finally, and perhaps most important, we can use the taxonomy to
develop typological theories about how combinations of mechanisms
interact in shaping outcomes for specified populations. A typological
theory is a theory that not only specifies individual independent variables and the hypothesized causal mechanisms that shape their effects,
but provides “contingent generalizations on how and under what
conditions they [these variables] behave in specified conjunctions or
configurations to produce effects on specified dependent variables”
(George and Bennett 2005, 235). Typological theories attempt to
address complex causal relations, including non-linear relations,
high-order interactions effects, and processes involving many variables. The taxonomy thus provides the building blocks of theorized
mechanisms that can be brought together in specified conjunctions to
develop typological theories on how combinations of variables behave.
The empirical chapters in this book illustrate each of these six kinds
of contributions. Generally, as most of these chapters deal with one
or a few cases, they collectively make more contributions in the first
three of these areas than in the last three (cross-case generalizations,
Causal mechanisms & typological theories
217
identification of theoretical scope conditions, and development of
typological theories). I address each of these actual and potential
contributions in turn.
This clarifies the mechanisms each author has studied and those they
have either set aside or perhaps overlooked. Nome and Weidmann,
for example, explicitly note that their agent-based modeling approach
necessarily sets aside structure-to-structure mechanisms or emergent
properties. Adamson indicates that her account de-emphasizes structural
factors without entirely ignoring them. More generally, the empirical
chapters emphasize agent-to-agent power/resource and legitimacy mechanisms, structure-to-agent power mechanisms, and agent-to-structure
legitimacy mechanisms; they give less attention to institutional efficiency
mechanisms and structure-to-structure mechanisms.
These relatively neglected mechanisms could have been used to address
more fully issues such as how particular rebel leaders and ideas won out
over more traditional leaders and concepts – as with Joseph Kony and the
Lord’s Resistance Army (LRA) in Uganda, the Kurdistan Workers’ Party
(PKK) in Germany, and the more radically Islamic rebel fighters in
Chechnya. In part, these leaders and their ideas likely prevailed because
they were functionally efficient in serving purposes for other actors, in
addition to being able to gain legitimacy and control resources.
Most of the codings in Table 8.2 are fairly straightforward, involving terms used directly by the authors, but some reflect more implicit
arguments. For example, Schmitz notes that lower-ranking rebels took
up the amnesty offered by the Ugandan government, while top rebel
leaders rejected the offer. He does not explicitly argue, however, that
this process may have created an “adverse selection” effect by leaving
the most committed and violent rebels in even greater control of the
LRA, nor does he suggest that this might help account for the increase
in violence that followed. Even though it works through individual
choices first by Ugandan government officials and then by rebel
leaders, the adverse selection going on here is a structure-to-structure
mechanism: the amnesty policy constitutes an institutional rule, the
outcome (a more violent LRA) is also a social structure, and that
outcome is presumably not intended by the Ugandan government.
Table 8.2 reveals that framing, resources, and brokerage received
considerable attention in the empirical chapters. Yet the chapters
devoted less attention to the functional efficiency row and the structureto-structure column. One exception is Hamberg’s consideration as an
Andrew Bennett
218
Table 8.2 Locating the causal mechanisms invoked in the empirical
chapters within the taxonomy
Structure
Agent to agent to agent
Legitimacy Emulation B
Socialization
A, B, S, HL
Learning B, S
Teaching B
Material
power
Adaptation
NW
Ethnic
bidding A
Shaming S
Agent to
structure
Structure to
structure
Norm
entrepreneurs
NW, S
Framing
A, B, H, S
Ethnic/leadership
bidding A, B
Agents change
Demographic
change A,
material
HL
power balance
A, B, HL
Hegemonic
Resources as
incentives
socialization
B, H, S
B, H, S
Resources
enable
A, B, HL, S
Networks/
brokerage
A, B, H,
HL, NW
Functional Emulation of
Evolutionary Functional
selection of
efficiency
successful
innovation B
agents B, S
efficient
leaders or
practices
A, B, HL
Key: A Adamson; B Bakke; H Hamberg; HL
S Schmitz; NW Nome and Weidmann
Adverse
selection S
Harpviken and Lischer;
alternative hypothesis whether the use of child soldiers became functionally less efficient for the Sudan People’s Liberation Army/Movement
(SPLA/M) than the use of adult soldiers. He concludes this was not a
strong explanation in his case. The structural opportunities for rebellion
emphasized by Fearon and Laitin (2003) get only passing references in the
chapters as well.
Future research might usefully explore these dimensions and their
underlying mechanisms more closely, addressing questions such as: do
Causal mechanisms & typological theories
219
considerations of functional efficiency help explain why certain practices and leaders rise and fall? How do demographic change, economic
change, and broad changes in information technologies and global
discourses (such as the growing prominence of discussions on the
“responsibility to protect”) shape the structural context within which
transnational actors operate? By seeking to assassinate leaders of
transnational movements that oppose them, have states unintentionally brought younger and more radical leaders to power?
Regarding the second way in which the taxonomy can be used, the
empirical chapters make several very helpful contributions by drilling
down and disaggregating theorized mechanisms in one or another of
its categories. Nome and Weidmann distinguish norm entrepreneurship from adaptation, emulation, and learning. Adamson differentiates
frame bridging (the linking of two different frames) and frame
extension (the extension of a frame to new issues or groups). Bakke
clarifies the differences among mediated, relational, and non-relational
diffusion (see also Wood, this volume).
The third kind of contribution – historical explanations of individual
cases using a variety of theories on causal mechanisms – is the most
readily identifiable in the volume’s empirical studies. It is also an
important form of knowledge accumulation that often gets overlooked
in those social sciences that prize theoretical generalizations across
cases over theoretical explanations of specific historical ones. With
the exception of the chapter by Nome and Weidmann, each of the
substantive chapters focuses upon an empirical puzzle and draws on a
variety of theorized mechanisms to explain it. Why did the SPLA/M,
in contrast to other rebellious groups, demobilize its child soldiers
(Hamberg)? Why did the LRA (Schmitz) and the Chechen rebels
(Bakke) shift the framing of their conflicts? How did Turks in
Germany who could not even speak Kurdish come to think of themselves as Kurds and contribute to the PKK (Adamson)? Why did
some refugees and not others engage in violence upon return to their
home countries (Harpviken and Lischer)? The authors then proceed to
explain these historical puzzles with the combinations of mechanisms
identified in Table 8.2.
A fourth possible contribution of the taxonomy – improving our
understanding and estimates of the magnitude of the causal effects of
the variables in play – is under-utilized, which is understandable given
that the qualitative methods used in most of the chapters do not allow
220
Andrew Bennett
estimates of causal effects in the way that regression analyses might.
Nonetheless, contributors, on several occasions, do give assessments of
which mechanisms might have greater, more lasting, or more robust
effects than others. In particular, Nome and Weidmann use agentbased modeling to derive the result that the mechanisms involved in
norm entrepreneurship are more powerful than adaptation mechanisms in the conditions they specify, including social complexity. Also,
Hamberg argues that “naming and shaming” has not proved sufficient
for the demobilization of child soldiers; indeed, he comes close to
suggesting that outside actors’ control of non-substitutable resources
that rebel groups desire is a necessary condition for them to demobilize
such soldiers. Similarly, Schmitz notes that the LRA’s learning of
reprehensible tactics in Sudan proved more powerful than the efforts
of states, NGOs, and the International Criminal Court to persuade it to
respect human rights.
The taxonomy’s fifth contribution – the specification of scope conditions – also receives little attention, with the empirical studies only
occasionally making claims regarding them, a point Wood (this
volume) has also noted. Still, the chapters do at times make brief
comparisons to other cases or use counterfactual analysis to suggest
possible scope conditions. Harpviken and Lischer, drawing on their
longitudinal analysis of refugees in Afghanistan and Rwanda, conclude
that groups deeply socialized into adopting certain norms and practices
will continue to use these to guide their behavior even when resources
decline. In contrast, groups that adopt norms for the instrumental
purpose of gaining resources will no longer follow these norms when
resources decline. The accounts by Adamson and Bakke suggest that
PKK militants and transnational jihadists would have had less success
in Germany and Chechnya, respectively, if local elites had more independent resources or if imported identities had not resonated with preexisting historical narratives and local actors’ lived experiences of
alienation.
This volume’s focus on individual cases limits efforts at the cumulative development of theories that might apply across them. The empirical chapters at times make reference to additional positive or negative
cases, but this is typically done to set the stage for the authors’ chosen
cases as particularly puzzling or interesting. An example is provided by
Hamberg’s analysis, where he claims the SPLA is the only successful
case of getting a rebel movement to demobilize child soldiers.
Causal mechanisms & typological theories
221
Bakke’s chapter arguably goes the farthest toward identifying generalizable and often structural factors that affect the ways in which
mobilization mechanisms play out. She notes that these factors include
the prevalence of conflict in the region, the strength of the target state
(weak states present low opportunity costs for insurgents, strong states
may push insurgents to seek havens and help across borders), the
presence or absence of ethnic or ideological ties between insurgents
and neighboring states, the existence or absence of brokers between
local actors and diasporas, and the willingness of local fighters to
accept foreign fighters and/or the (in)ability of locals to keep out
foreign fighters. Still, neither Bakke nor the other authors move farther
toward cumulative theorizing by identifying the population of positive
and negative cases relevant to the phenomenon they study or discussing more than a few comparative cases.
An additional reason the empirical chapters do not attempt more
ambitious cross-case generalizations is that the field has only recently
begun to develop new ways of couching them, ones that allow for
complex interactions among variables. Statistical analyses seldom
model interaction effects beyond two variables. At the same time,
qualitative analyses of one or a few cases, like those in this volume,
rarely attempt to provide an empirical basis or a theoretical framework
for generalizations, even contingent ones.
Case studies, causal mechanisms, and typological theories
The foregoing discussion brings us directly to the sixth type of contribution: using the taxonomy of mechanisms to help build typological
theories. Constructing a typological theory involves both deductive
and inductive reasoning. Deductively, the analyst defines the dependent variable of interest and uses prior theories – such as those included
in the taxonomy – to identify the relevant independent variables. The
analyst then creates a typological space (or what is known in logic as a
“property space”) that consists of all the possible combinations of the
variables. Inductively, the analyst begins placing known cases into the
typological space according to preliminary knowledge of the values of
the variables in those cases. The analyst can then iterate between what
was theorized a priori, what is known empirically, and what is learned
from additional empirical study, refining the typological space and
possibly re-conceptualizing variables to higher or lower levels of
222
Andrew Bennett
aggregation depending on the complexity of the space and the analyst’s
research goals (George and Bennett 2005, chapter 11; Elman 2005).
A study I did with colleagues on alliance burden-sharing in the
American-led Desert Storm coalition in the 1991 Gulf War (Bennett,
Lepgold, and Unger 1994) illustrates the potential value of typological
theorizing. Most research on alliances up to that time had focused on
only one or a few mechanisms that led to alliance contributions,
including collective action dynamics, the domestic politics of potential
allies, the security threats to potential allies, or the alliance security
dilemma. In contrast, we developed a typological theory on how
combinations of these variables might lead to or inhibit alliance contributions. We then briefly examined some states that contributed and
other states that did not contribute to the Gulf War coalition, using this
to refine our theory. We next selected a smaller number of cases for
close study and process tracing to test our theory against more detailed
evidence from the cases. At this point, even though we had partly
constructed our theory on the basis of preliminary knowledge of the
outcomes of the cases, our theory could still have proved to be wrong
about the processes through which those outcomes arose.
One of the most striking findings, backed up by both the outcomes
and the process-tracing evidence, was that Germany and Japan each
contributed over $10 billion to the 1991 coalition because of the
alliance security dilemma (in these instances, security dependence on
the United States). This was true even though the variables for all the
other three theories incorporated into our typological account weighed
against Germany and Japan making large contributions. This finding
limited the scope conditions of collective action theory, which had
dominated the alliance literature and which wrongly predicted free
riding by Germany and Japan. The typological theory we developed
provided a useful framework that other scholars have built upon and
applied to additional coalitions (Baltrusaitis 2009; Davidson 2011).
Like alliance burden-sharing, the transnational dimension of civil
war is an enormously complex subject, and present space permits only
a first cut at a typological theory on this topic. Even this preliminary
effort, however, highlights the costs and benefits of typological
theorizing.
To build an illustrative typological theory on transnational actors
and civil conflicts, I examine whether rebel groups have external
sanctuaries, as this is a critical factor in two ways. First, rebel
Causal mechanisms & typological theories
223
sanctuaries in neighboring states are common, having played a role in
more than half of the insurgencies since 1945 (Checkel, this volume;
Salehyan 2009). Second, the presence or absence of cross-border
sanctuaries strongly shapes the fighting and bargaining among the
rebels’ home state government, the rebels’ host state, the international
diaspora, third countries, NGOs, and rebel leaders. As Salehyan
(2009) points out, cross-border havens lower the opportunity costs
for rebels to mobilize, but they also complicate the bargaining problem
between rebels and their home governments by intensifying
information asymmetries and bringing an additional party – the host
government – into the bargaining process. The presence or absence of
cross-border rebel havens thus powerfully shapes the relationship
between transnational actors and civil conflicts, and rebel havens are
accordingly an important factor in the chapters in this volume by
Adamson, Bakke, Hamberg, Harpviken and Lischer, and Schmitz.
To develop a typological theory on cross-border havens, I introduce
three independent variables. The first of these concerns whether rebels
control territory in their home state. If they do control home state
territory, this gives them greater bargaining leverage vis-à-vis both
the home state government and potential host governments.
A second variable is whether the government of a potential host
state, a minority group in that state, or neither, supports the rebels’
agenda. Government support can provide resources and a haven, while
support from a minority group can provide a haven but can also put
rebels at odds with the host state government if it fears any growth in
the minority group’s power. A loss of host government support can
force rebel leaders to seek support from wider diaspora communities,
third governments, NGOs, or homeland groups, and can force a return
of the diaspora to its homeland. This is evident in the ups and downs of
the various rebel-host relations documented in this volume: the LRA
and Sudan; the Taliban and Pakistan; the Tutsis and Uganda; and the
SPLA/M and Uganda.
A third variable is whether the potential host is a strong or weak
state. If the host country has a strong state, it may use its resources to
manipulate the rebels to serve the host’s goals, as happened with the
LRA in Sudan and the Taliban in Pakistan. If the host country has a
weak state, the rebels will be able to pursue their own goals, and they
could even become influential in the politics of the host state, as
happened for a time with the Tutsis in Uganda.
224
Andrew Bennett
The first of these independent variables (presence or absence of a
rebel haven in the home state) is dichotomous, as is the third independent variable (strength or weakness of the host state). The second
variable, support for rebels from the host government, a host state
minority, or neither, is trichotomous. There are thus 223 or 12
possible combinations of these independent variables, or 12 types, in
the typological space.
I define the dependent variable as the nature of the triangular relations and bargaining among the home government, the host government, and the rebels (for a similar approach, see Salehyan 2009). If
rebels control territory in their home state, they will have more
bargaining leverage and autonomy vis-à-vis both their home government and potential host governments. This presence or absence of a
rebel haven in the home state then interacts with the six combinations
that are possible between the second and third independent variables.
These six combinations represent all of the possible mixes of three
kinds of host state support for the rebels’ agenda (government, minority, or neither supports the rebels) and two levels of host state strength
(the state is either strong enough to expel rebels if it so chooses, or too
weak to do so). I provide descriptive labels for each of these six
combinations as follows:
1. A “meddlesome neighbor” arises when a strong host state supports
the rebels against the home state.
2. A “cooperative neighbor” is when a strong host state opposes the
rebels but needs home state help to keep them out because a
minority in the host state helps the rebels.
3. No haven exists when the neighboring state is strong and neither
that state nor a strong minority within it supports the rebels.
4. A “weak troublemaker” is when a weak host state backs the rebels.
5. An “eager junior partner” exists when a weak host state needs help
from the home state to fight the rebels and their host state minority
supporters.
6. A “weak partner” is when a weak state is unable to keep out rebels
even though the rebels lack support from a minority in the host
state.
Having developed the independent and dependent variables, the next
step is to combine them into a typological space comprising all 12
possible combinations of these variables. This typological space is
Causal mechanisms & typological theories
225
outlined in Table 8.3. Table 8.3 also represents the next step in the
development of the typological theory: the categorization of the extant
cases in the typological space based on a preliminary understanding of
the values of the variables in each case. The table thus represents the
typological theory sketched out above and places within the resulting
typological space all the cases discussed in the chapters by Adamson,
Bakke, Hamberg, Harpviken and Lischer, and Schmitz. I categorized
the cases based on my reading of the chapters, rather than doing so
based on dedicated measurements by the authors themselves or any
expert independent knowledge of my own. In addition, some cases
defy easy measurement, as I have simplified the typological theory by
using dichotomous measures of variables that are in fact continuous.
The case of Pashtun refugees in Pakistan, for example, is difficult to
characterize in view of ambiguities in the Pakistani government’s policies toward the Pashtun-dominated Taliban. The goal of the present
exercise is to illustrate the potential value of typological theorizing
rather than to categorize the cases definitively. Indeed, part of the value
of such theorizing is that by making the categorization of cases explicit,
it sparks useful discussions on the underlying concepts and the operationalization and measurement of the variables.
Four contributions flow from my effort at bringing together variables from different theories about causal mechanisms to construct this
typological theory. First, it helps us see which cases should have similar
processes and outcomes. According to Table 8.3, the Chechen rebels
and the SPLA/M are the same type of cases, and they do indeed have
arguably similar processes and outcomes. The Chechen rebels, lacking
a secure home base or cross-border haven, were heavily reliant on the
aid of transnational actors – in this case, a transnational radical Islamic
movement that provided weapons, funds, and training in exchange for
the Chechens’ adoption of a more radical Islamic agenda. As Bakke
notes, the Chechen case might have turned out differently if local
leaders had been better able to push themes of nationalism and independence over those of jihad and sharia. However, the secular leaders
lacked resources to win the framing fight. The other case in this type,
the SPLA/M group in Sudan, displayed a similar dynamic. Hamberg
recounts how the SPLA/M, lacking a secure base at home or abroad,
was not convinced by normative pressures from NGOs to stop using
child soldiers, but proved very open to US pressure to demobilize child
soldiers in exchange for US material support.
Table 8.3 A typological theory on triangular bargaining among rebels, home states, and host states
Cases:
Home/host, dates
Tutsis in Rwanda 1990 present
Hutus in Rwanda, 1996 present
LRA/Sudan, 1990 2002
LRA/Democratic Republic of
Congo, 2002 2005
Chechen Rebels 1999 present
SPLA/M
Tutsis/Uganda, 1961 late 1980s
Hutus/Zaire, 1994 1996
Pashtuns/Pakistan 1979 1989
Pashtuns/Pakistan 2001 present
Turkish Kurds/Iraq 1991 present
LRA/Central African Republic
2005 present
Rebels hold
territory in
home state?
Host government, host
minority, or neither
support rebels (G, M, N)
Host state
strong or
weak?
Y
G
S
Y
M
S
Y
N
S
Y
G
W
Weak troublemaker, rebels have
autonomy
Y
Y
N
M
N
G
W
W
S
N
M
S
N
N
S
Eager junior partner, rebels have leverage
Weak partner, rebels have autonomy
Meddlesome neighbor, rebels dependent
on host
Cooperative neighbor, rebels depend on
minority in host
No haven; rebels vs. home government
N
G
W
Weak troublemaker, rebels depend on
host
N
M
W
Eager junior partner, rebels autonomous
N
N
W
Weak partner, rebels have autonomy but
are vulnerable
Outcome: predicted nature of the
triangular bargaining
Meddlesome neighbor, rebels have
leverage
Cooperative neighbor, rebels have
leverage
No haven; rebels vs. home government
Causal mechanisms & typological theories
227
Another set of parallels concerns the fates of the Rwandan Hutus
and Tutsis. For some years in the 1960s through the 1980s, the Tutsis
had a haven in Uganda. Similarly, the Hutus fled in 1994 to a haven in
Zaire. While the governments of Uganda and Zaire were weak and
willing to support or at least tolerate refugees from Rwanda, the
militants among these refugees had relative autonomy. Harpviken
and Lischer recount, for example, how Ugandan leader Yoweri
Museveni drew on Tutsi support to take power in 1986. The Tutsis
lost their leverage, however, when Museveni turned on the Tutsis in the
late 1980s and forced them back into Rwanda. When the Tutsis lost
their haven in Uganda, and when the Hutus later lost their haven in
Zaire, each group was forced back into Rwanda, where the groups
entered into zero-sum conflicts with one another as they lacked any
cross-border haven into which they could retreat.
If cases in the same type have different outcomes from one another,
or if any cases do not fit the predictions of the typological theory, these
merit closer study to see if the outcomes might be due to variables
omitted from the theoretical framework. There are no obvious anomalies in Table 8.3, but categorizing additional cases and placing them
within the typological space could reveal some anomalies worth
exploring.
Second, the typological theory helps identify “most similar cases”
that differ in one independent variable and in the dependent variable,
and that can thus be usefully compared. This allows the use of process
tracing to help assess whether the independent variable that differs
between the cases accounts for the difference in their outcomes. There
are many pairs of cases in Table 8.3 that differ in one independent
variable. The cases that have no base in the home state and have
minority support in a weak neighboring state might be compared to
those that lack a base in the home state but have support from a weak
neighboring government (i.e., the NMW cases compared to the NGW
cases). Cases lacking a home base but having the support of a weak
cross-border government could also be compared to those that have
both a home base and foreign support from a weak government.
Third, the theory facilitates the explanation of individual cases.
Schmitz notes, for example, that the LRA’s Joseph Kony avoids reliance on outsiders. However, he does not note that Kony’s autonomy
was in part an outcome of his structural position. Being able to retreat
to either his base in Uganda or his cross-border bases made Kony
228
Andrew Bennett
autonomous from 1990 to 2005. In contrast, Kony’s loss of crossborder bases since 2005 and the weakening of his influence inside
Uganda have undercut his bargaining power vis-à-vis the Ugandan
government and made him vulnerable. Similarly, although this barebones typological theory omits how rebel leaders framed their messages, adding this variable to the framework would provide a sharper
focus on whether the changes in framing by Kony, the Kurds in
Turkey, and the Pashtuns in Afghanistan were driven more by structural needs and opportunities or by leaders’ ideological predilections.
Pashtun refugees in Pakistan, for example, appear to have tempered
their support for an independent Pashtun state in order to win support
from or at least moderate the opposition of the Pakistani government.
Fourth, the typological theory advanced here provides an initial
framework that scholars can expand by adding new independent
variables. Most of the variables in this typological theory are drawn
from structural theories about power mechanisms, but the theory
could be expanded, albeit with a loss of parsimony, by incorporating
additional structural or agent-centered variables. Additional structural
variables might include the host state’s international autonomy or
dependence on other states with an interest in the conflict, the presence
or absence of lootable resources, geographic terrain that provides rebel
havens, rebels’ social networks and legitimacy among potential followers and allies, and military and spying technologies that raise or lower
the relative costs of government repression and rebel mobilization.
Agent-centered variables could include the frames that rebel leaders
choose, as well as their skills at negotiations and the use of force.
Adding variables of course adds to the complexity of the theory, but
researchers can pare back this complexity by controlling for some of
the variables, thereby exploring only subsets of the full typological
space in any one study.
Conclusions
The field of political science has moved increasingly over the last two
decades toward mechanism-based explanations of complex phenomena. This shift is related to the development of variants of scientific
realism in the philosophy of science, but it has been hampered by
limited understanding among political scientists of how mechanismbased explanations differ from explanations built upon earlier
Causal mechanisms & typological theories
229
philosophies of science. As Peter Hall has persuasively argued (2003),
our ontological theories of how politics work, which increasingly
embrace complexity, have outrun our epistemological notions of how
to study politics, which still cling to the vestiges of forms of positivism
that lost favor among philosophers decades ago.
A focus on explanation via reference to causal mechanisms offers
one way of bringing our ontological assumptions, epistemological
approaches, and research methods back into alignment. Yet political
scientists have been hesitant to commit fully to this move because they
have lacked a clear sense of the philosophical costs and benefits of
mechanism-based explanations. The present chapter has argued that
although mechanism-oriented explanations are not without their own
drawbacks, they are an improvement over Kuhn’s concept of
“paradigms” and Lakatos’ notion of “research programs.” Whereas
Kuhnian paradigms and Lakatosian research programs both foundered, in different ways, on the difficulties of justifying large sets of
partially testable interrelated ideas, the concept of theories about discrete causal mechanisms allows for middle-range theories that are more
localized, if also, more complex.
At the same time, political scientists need assurance that explanation
via mechanisms does not entirely lack the key attraction of paradigms
or research programs: a structured discourse that provides a framework around which we can organize cumulative research findings.
Why should we move away from the “isms” – realism, liberalism,
and constructivism in the IR subfield; rational choice, historical
institutionalism and other “isms” in the study of American and comparative politics – and toward causal mechanisms if the latter contribute only to a hodgepodge of discrete explanations of individual cases?
Here, the taxonomy of causal mechanisms introduced above shows
how extant paradigms and research programs have implicitly relied on
causal mechanisms all along and can be mapped onto an approach that
focuses on explanatory mechanisms without reifying them into grand
schools of thought.
This volume’s empirical chapters demonstrate how the taxonomy
allows cumulative research along many dimensions: identification of
the mechanisms invoked and overlooked in particular explanations;
differentiation of subtypes of mechanisms; explanations of historical
cases using a variety of theorized mechanisms; assessment of the relative strength of alternative mechanisms in cases; delineation of the
230
Andrew Bennett
scope conditions under which theorized mechanisms operate; theories
that cut across small groups of cases; and, finally, typological theories
that provide contingent generalizations about populations of cases.
As the empirical studies largely focus on one or a few conflicts, the
first three of these contributions are stronger than the last three.
Yet even here it is possible to demonstrate the potential contributions
of constructing a typological theory of civil conflict, which examines
how combinations of mechanisms interact in different structural
circumstances and among mixes of actors with varying endowments
of power, legitimacy, and institutional access.
Mechanism-oriented theorizing poses important costs, particularly a
loss of parsimony compared to extant paradigms and research programs. Still, researchers using this approach to theory-building can
choose different tradeoffs along the spectrum between parsimony and
complexity. In the end, there is a strong philosophical basis for rooting
the study of politics in theories about causal mechanisms. More
important, it is possible to do so without losing a structured discourse
and cumulation of research findings – as this book’s cases and their
extension suggest (see also Wood, this volume).
|
9
Transnational dynamics of civil war:
where do we go from here?
elisabeth jean wood
Introduction
The transnational aspects of conflict are a pervasive theme in news
reports. Revolutionary uprisings spread from Tunisia to many countries
of the Middle East and North Africa in 2011, led in many cases by
activists who had trained with activists from other countries. Afghan
insurgents retreat across the Pakistan border to relative safe havens.
Kenyan troops cross into Somalia to fight against the al-Shabab insurgency, itself responsible for attacks in Kenya, where some 500,000
Somali refugees have sought refuge from drought and violence. Transnational drug supply routes finance non-state actors in Colombia,
both rebels and narco-paramilitary groups. US drones target members
of al-Qaeda in Yemen, including US citizens – themselves transnational
insurgents – as well as in Pakistan. Many Tamil emigrants from
Sri Lanka contributed regularly to the Liberation Tigers of Tamil Eelam
(LTTE), often under the threat of violence from LTTE sympathizers
in the diaspora.
Although these examples may suggest that the transnational dimension of civil conflict is increasing, transnational flows of various kinds
have supported actors engaged in conflict since well before the end of
the Cold War. In its struggle against apartheid, the African National
Congress had a network of offices across Europe and training camps
in a number of African countries. Both leftist insurgent groups and
counterinsurgent states drew on transnational ties during the 1970s
and 1980s in Latin America. Revolutionary uprisings spread across
Eastern Europe and the states of the former Soviet Union beginning in
the late 1980s, and revolution spread from France to several European
and Latin American countries in 1848.
Despite its evident importance in both contemporary and historical
cases, scholars have only recently analyzed systematically the role
of the transnational in civil conflict. The reasons for the neglect are
231
232
Elisabeth Jean Wood
diverse. One reason was the division of labor between comparative
politics scholars, who tended to focus on domestic politics at
the expense of international influences, and international relations
scholars whose work on transnational politics focused on the diffusion
of international norms or economic policies but paid little attention
to transnational aspects of civil conflict (see also Checkel, this volume).
Another challenge that quantitative scholars confronted was the
difficulty of building databases that adequately capture transnational
influences, often covert. Scholars immersed in single case studies or
regional comparisons often shrank from framing their findings as
generalizations that could be tested in other settings.
Beginning less than a decade ago, scholars have increasingly documented the role of transnational dynamics on civil conflict – on conflict
onset, duration, and sometimes its attenuation. The literature is
diverse, ranging from quantitative analysis of new data sets that
include variables for refugees, insurgent bases, transnational ethnic
groups, various forms of intervention by states, to careful qualitative
analysis of individual cases of transnational politics undergirding
civil conflict. The chapters in this volume are exemplary works of the
latter type, pushing forward the frontiers of our understanding of
transnational conflict dynamics through their identification of relevant
causal mechanisms, their careful assessment of the evidence for those
mechanisms as well as rival explanations (in some cases including
original data gathered through field research), and their articulation
(sometimes implicit) of the conditions under which those mechanisms
are likely to shape conflict and its aftermath.
In this chapter, I assess the state of the literature on transnational
dynamics of civil conflict with particular attention to the contributions
of this volume. I also discuss the volume’s methodological advances
and challenges, focusing on process tracing as a technique to identify
causal mechanisms and to demonstrate their effect. I begin with a
discussion of the recent literature on the transnational diffusion of
policy, as it in some ways sets the standard for the study of diffusion.
That discussion also lays the groundwork for an assessment of the
contributions of this volume to our understanding of conflict as well
as to the broader literature. I emphasize the recent move to study
how diffusion of a policy depends on the partisan composition of the
government considering its adoption, as well as on its political (not
merely policy) consequences where it was adopted earlier.
Where do we go from here?
233
I then review the recent quantitative literature, similarly emphasizing
work that shows how the risk of conflict depends on the ethnic
composition of the state (in particular, whether large ethnic groups
are excluded from political power) as well as of its neighbors. Next,
I assess this volume’s contributions to our understanding of civil
conflict, diffusion, and qualitative methods. Finally, in the conclusion,
I identify promising research topics arising from this volume.
Diffusion: a metaphor, an outcome, a process?
Imagine that the trees spread out across a broad swathe of rolling
grassland all shriveled up and quickly died, with those further to
the south dying first and those to the north last. We might say that a
drought had spread, or diffused, across the area from south to north,
as evidenced by the sequence and location of dead trees. What precisely
would be the causal mechanisms behind the diffusion of drought?
It might be the case that the trees died independently as rains failed
first in the south and later in the north. Tree death “diffused” across
the landscape, but in this case, the “diffusion” of the drought is a
metaphor as there was in fact no interdependence between the trees
associated with their death. Or perhaps the trees died because a pest
spread from tree to tree beginning in the south, killing trees as it went.
In this case, the material cause of tree death itself diffused across
tree borders.
A more difficult case would be if a very water-intensive crop was
planted beyond the southern border, depleting underground water
there. The roots of the trees along the southern border tried to make
up for decreased water supply by drawing water from under their
neighbors just to the north, depleting some of their water. Each group
of trees tried to adjust by taking water from farther north, with the
result that all gradually shriveled as none could get enough. In this
case, an absence of sufficient water spread as in the first case, but there
is an interdependence between the trees that causes the drought to
spread as in the second case.
I begin this way to contrast diffusion as a mere metaphor – in which
some outcome “spreads” but without interdependence between the
units – with diffusion as a process – in which the outcome in one
unit occurs because it occurred in another: they are interdependent in
some way. Is a geographical cluster of conflict-riven states a result of
234
Elisabeth Jean Wood
diffusion, or is it merely the case that domestic factors that make conflict
more likely were equally high across the region with no “spillover”
between one conflict to the next? To distinguish the diffusion of an
outcome from its mere spread, it is necessary to specify the causal
mechanism whereby the interdependence of units results in similarity
of outcome.1 In this section, I illustrate these challenges through a
discussion of policy diffusion, the social science literature that has
most explicitly addressed them.
The causal mechanisms underlying diffusion of policy are generally
understood as falling into four categories: emulation, learning,
coercion, and competition (Dobbin et al. 2007). In emulation, policymakers adopt a policy innovation not because they believe it the best for
their country but because they strive to emulate early adopters or
take the advice of policy experts, thereby signaling their conformity,
modernity, or innovator status. Interdependence in this instance is
socially constructed: policy-makers willingly emulate global policy
innovators (or perhaps they emulate innovations among their
peers) with no consideration for whether the policy is appropriate for
their polity.
In learning, policy-makers evaluate the new policy – its likely costs
and benefits in their setting – and adopt it as the policy more likely to
bring about the desired outcome. It is the assessed efficacy of the policy
rather than mere copying that distinguishes learning from emulation.
Learning may be rational in that the policy-maker carefully evaluates
all sources and options for the polity’s own circumstances (and prior
beliefs are not so strong as to undermine Bayesian updating, Meseguer
2006). Or it may be boundedly rational in which some sources are
weighed more heavily than others for reasons that are not rational.
Chile’s pension privatization policy, for example, was copied across
much of the Southern Cone, despite differences in welfare systems
(Weyland 2007).
Similarly, revolution diffused across much of Europe and Latin
America in 1848 through the mechanism of boundedly rational
learning, according to Weyland (2009): opposition members across
many countries learned from the French example that revolution was
possible but did not adequately account for their distinct circumstances
1
For a helpful formulation of varying degrees of interdependence, see Franzese and
Hays (2008, 752 754).
Where do we go from here?
235
(except in extreme cases, which showed that learning was rational,
not merely emulation). Jeff Checkel (2001) points out that a stronger
form of learning – “complex learning” – occurs when policy-makers
come to be persuaded through principled debate; indeed, in some cases
learning may lead to the development of new interests, norms, or
identities rather than merely a more informed cost benefit calculation.
In coercion, policy-makers face incentives that compel them to adopt
the new policy because their choice has been limited by the conditions
imposed by strong states. Or coercive power takes the form of promoting hegemonic ideas such that deviation is difficult to imagine.
In competition, policy-makers adopt a policy innovation because their
rivals have done so (or will do so soon) and thus they must as well to
remain competitive. As in coercion, policy-makers respond to incentives
but choice is not constrained in principle, despite there being only one
outcome – adoption – that is rational in the competitive circumstances.
Bearing in mind the example of the pest spreading across the grasslands to infect trees, migration is a fifth category that is very relevant
for transnational conflict dynamics, as we will see: a policy is adopted
in a new jurisdiction because an agent moves to the new polity and
carries the policy along (Franzese and Hays 2008).
Despite the articulation of distinct causal mechanisms underlying
diffusion, in much of the literature, the causal mechanism is theorized
and the spread of an outcome is demonstrated, but without showing
that the particular mechanism was, in fact, the operative cause (Dobbin
et al. 2007, 463). Gilardi (2010, 650) emphasizes the accumulating
evidence that policies do in fact diffuse, but goes on to argue:
On the other hand, the literature has been less successful in unpacking
diffusion empirically, that is, in identifying specific diffusion mechanisms.
Policies diffuse, but why? There is agreement that competition, learning, and
social emulation are the main drivers of diffusion, but empirical evidence
usually is ambiguous and unable to discriminate convincingly among these
different explanations. Learning has been a particularly elusive hypothesis.
While more recent literature attempts to show that a particular diffusion
mechanism is at work, interpretation is complicated by the overlap
between mechanisms. In both emulation and bounded learning, for
example, policy-makers adopt policies that may be inappropriate for
their polity, but do so for different reasons that may be difficult to
observe (Meseguer and Gilardi 2009, 530–531).
236
Elisabeth Jean Wood
The challenges are theoretical, conceptual, and methodological.
Meseguer and Gilardi (2009, 531–533; see also Gilardi 2010) point
out that the literature makes broadly homogenizing assumptions
that all actors respond in the same way to the same things, implying
an underlying theory of diffusion in which the mechanisms (each itself
sufficient) are additive causal factors. Moreover, they argue, it neglects
the possibility that policy-makers assess the likely political consequences of policies, as well as their likely outcome. The underlying
problem is the absence of a theory of diffusion with hypotheses
specifying the conditions under which each mechanism is likely to be
causally important. Methodologically, if the adoption of some policy
in country B after witnessing its adoption in country A depends on
particular characteristics of country A and/or B, standard (non-dyadic)
quantitative approaches to analyzing diffusion from A to B will likely
fail to isolate the relevant, dyadic effect (Franzese and Hays 2008;
Gilardi 2010).
Gilardi’s “Who learns from what in the policy diffusion process?”
(2010) addresses several of these concerns, highlighting the salience of
political consequences as well as policy outcomes in a model in which
policy-makers learn from others but what they learn is conditional on
their ideology. Policy-makers are not equally sensitive to new information about the likely effects of policy change, he argues, and they
learn from both policy and political consequences of reform. Prior
beliefs (which reflect ideological positions) shape the interpretation of
new evidence, and therefore differently positioned policy-makers react
differently to the same evidence. In his analysis of unemployment
benefits (specifically, the replacement rate, which is the share of the
salary a worker receives through unemployment insurance after losing
his or her job) in 18 OECD countries, Gilardi demonstrates the importance of the inclusion of these political considerations: policy and
political outcomes in other countries do not affect policy decisions
per se, but do so in interaction with the partisan composition of
the government (2010, 658). Thus the proper question is “Who learns
from what . . .?” (the article’s title), and learning is more complicated
than usually recognized (2010, 661).
What is “learned” may thus be distinct from what was taught,
depending on conditions in the “learning” country. Acharya (2004)
argues that norm diffusion involves the localization of the norm, by
which he means the adaptation of the norm so that it is sufficiently
Where do we go from here?
237
congruent with local norms (see also Checkel 1999). The degree of
localization may vary across countries, across issues within the same
country, and also over time as when a norm’s initial acceptance (as a
localized norm) leads to the later displacement of the local norm by
the global norm. Based on field research in four countries, a team of
researchers lead by Peggy Levitt and Sally Merry (2009) found that
global ideas about women’s rights were appropriated and deployed
very differently in Peru, China, India, and the US. Localization or what
they term vernacularization – the process of appropriating and locally
adopting globally generated ideas and strategies – thus differs across
countries. Vernaculizers face two dilemmas, the need to resonate
with existing culture but also to advocate effectively for change. The
appearance of macro similarity in the adoption of a global “package”
of ideas and norms masks differences in interpretation on the ground.
Thus, researchers analyzing the patterns and consequences of policy
diffusion – even in the case of their formal adoption by countries with
stable institutions – face significant theoretical, methodological, and
empirical challenges. Policy scholars have nonetheless advanced this
research agenda through a mix of careful qualitative studies that
isolate particular mechanisms (Checkel 2001; Weyland 2007, 2009);
conceptual clarification that political consequences as well as policy
outcomes may diffuse, conditional on the politics of the policy-maker
(Gilardi 2010); the development of methods to capture such conditional diffusion; and the construction of new data sets.
Creating adequate conceptual tools, methodological approaches,
and data to address the challenges discussed above is more demanding
still for scholars studying the transnational dynamics of civil conflict.
In such settings, actors and processes may be covert, data often unavailable or imprecise, institutions unstable and contested, interviews
with knowledgeable insiders possible only in some settings (and often
only for one party to the conflict), and dissimulation of goals and tactics
a strategic imperative for many actors.
Transnational dynamics of civil conflict: recent quantitative
literature
Despite these difficulties, the quantitative literature on civil wars
has recently taken up the challenge of analyzing transnational conflict
dynamics. While the importance of transnational actors and spillover
238
Elisabeth Jean Wood
effects was a prominent theme in earlier qualitative literature, quantitative analysis of the significance and substantive effect of transnational
factors has surged in the past half decade or so (see also Checkel, this
volume). The dependent variable in this literature is usually the risk of
conflict onset, conflict duration, or the risk of renewed conflict after
war’s end. Similarly to the policy diffusion literature, advances have
depended on the development of new data sets that capture key variables
at the relevant unit of analysis and statistical approaches that look at
interdependence through a dyadic modeling lens. Moreover, just as the
diffusion of policy depends on the partisan composition of the government, many of the findings demonstrate the importance of the ethnic
configuration (not ethnic diversity) of the state and its neighbors in
transnational dynamics.
Transnational influences are significant in conflict onset, according
to a variety of studies. In their exhaustive exploration of the sensitivity
of empirical analysis of civil war onset to variation in data set and
model specification, Hegre and Sambanis (2006) found that contiguity
to a neighboring state in conflict was a robust predictor of conflict.
Gleditsch (2007) found that the effect on the risk of civil conflict onset
due to transnational factors was at least as large as domestic factors.
Salehyan (2007, 238–239) showed that civil war in a neighboring
country increases the risk of conflict onset by 85 percent. Moreover,
civil wars with outside involvement tend to last longer, result in more
deaths, and are more fatal (Salehyan, Gleditsch, and Cunningham
2011, 710).
The empirical demonstration that civil conflict spreads does not,
however, show that it diffuses. The observed regional clustering of
conflict might be due simply to a regional clustering of domestic factors
that enhance the risk of conflict. In these studies, the authors controlled
for domestic factors and thus the finding is that conflict, in fact,
diffuses rather than merely spreads.
What then are the mechanisms of transnational diffusion of conflict?
In principle, any of the mechanisms discussed above may underlie
the diffusion of conflict. Prospective insurgents may emulate those in
another country (whether or not the circumstances are propitious);
they may assess the costs and benefits of embarking on a similar
struggle and learn that the circumstances are propitious; they may be
coerced into rebellion (perhaps because of threats to family members
across the border). Incentives for conflict may increase because conflict
Where do we go from here?
239
across the border sharpens perception of grievances at home or
because economic externalities (decreased trade or growth) weaken
the state. Citizens or insurgents themselves may cross borders, bringing
with them grievances and perhaps organizational and technological
skills that deepen conflict in the host state.2
Although demonstrating that a particular mechanism is responsible
for conflict diffusion is challenging compared to doing so for policy
diffusion, scholars have nonetheless made several advances in systematically analyzing cross-national patterns.
Ethnic links across state borders have received perhaps the most
sustained scholarly attention. Gleditsch (2007) found that such links
predicted that conflict in one state raised the risk of conflict in neighbors that were also home to the ethnic group, and raised it more the
less democratic the neighbor and the less the integration of trade
between the countries. Buhaug and Gleditsch (2008) refined this finding
to show that transnational ethnic ties between a group in a separatist
conflict in one contributed to the emergence of a separatist conflict
in the second.
The nature of the causal mechanism underlying such “ethnic links”
was not, however, well specified. The existence of ethnic links indicated the presence of an ethnic group that spans both sides of a border
between one state in conflict and another whose increased risk of
conflict is the subject of study. Yet conflict erupts in few states compared
to the number of states that share an ethnic group with another. For
example, Russians in the many states that emerged after the end of the
Soviet Union did not rebel against their new governments (Cederman,
Gleditsch, Salehyan, and Wucherpfennig, forthcoming).
More recent work has built on the literature assessing the risk of
conflict as a result of the domestic configuration of ethnic groups, with
results that clearly establish the importance of ethnic politics for civil
conflict. The key move is threefold. First, research designs are dyadic,
focusing on the interaction between the challenger (a rebel organization in some cases, an ethnic group in others) and the state, with key
variables coded at the group level. This is an essential step away from
the mismatch between theories based on analysis of relations between a
challenger and the state and empirical tests that relied on inappropriate
2
Salehyan (2008) found that 55 percent of rebel groups conduct operations in
other states.
240
Elisabeth Jean Wood
country-level aggregate measures (Buhaug and Rod 2006; Cunningham,
Gleditsch, and Saleyhan 2009; Blattman and Miguel 2010; Cederman,
Wimmer, and Min 2010).
Second, analysis of ethnic conflict does not focus on ethnicity per
se (for example, ethnic diversity as measured by linguistic diversity) but
on the ethnic configuration of power, which includes measures for
the degree of political inclusion for all politically relevant ethnic
groups. This has now been compiled in an “Ethnic Power Relations”
(EPR) database (Cederman, Wimmer, and Min 2010). Third, analysis
of conflict increasingly relies on measures of the attributes of rebel
groups, rather than inadequate proxies, drawing on the “Non-State
Actor” (NSA) data set (Cunningham, Gleditsch, and Salehyan 2009).
In a seminal contribution based on the EPR data set, Cederman,
Wimmer, and Min (2010) show that the risk of conflict between a
particular ethnic group and the state increases with the degree of the
group’s exclusion from state power (especially if it experienced a loss
of power in the recent past), its size, and the experience of past conflict.
The effect is significant and substantial, and relevant in about half
of post-World War II conflicts. Thus it is not ethnic diversity per se
but exclusion from power that leads to conflict. Wucherpfennig,
Metternich, Cederman, and Gleditsch (2012) linked rebel groups in
the NSA data set with ethnic groups in the EPR data sets and found
that civil wars in which rebel organizations recruit from and fight on
behalf of excluded ethnic groups last longer than other civil conflicts.
Ethnic exclusion leads to strong collective solidarity (in part through
everyday humiliation of the excluded) and therefore high tolerance for
risk on the part of combatants, and greater legitimacy among civilians.
The state is reluctant to compromise, both because of its perceived
ethnic superiority and the fear that compromise would lead other
excluded groups to rebel.
As these studies have reinstated the importance of ethnicity in
the domestic political dynamics that lead to civil war, others have
established the salience of transnational ethnic links for the emergence
of conflict. In a dyadic analysis of states and ethnic groups in Eurasia
and North Africa, Cederman, Girardin, and Gleditsch (2009) found
that while the mere presence of ethnic kin in a neighboring state had no
significant effect on the risk of conflict (even when those kin were in
power), the probability of conflict rose sharply (a large and significant
effect) when the excluded minority was large (in the first state).
Where do we go from here?
241
However, this finding draws upon the coding of ethnic groups in the
Atlas Narodov, a coding of ethnic groups based only on language (not,
for example, religion) in the 1960s that includes politically irrelevant
groups and is thus a poor proxy for conflict-relevant ethnicity.
In a forthcoming paper, Cederman, Gleditsch, Salehyan, and
Wucherpfennig use the EPR data set with a new transnational extension coding the existence (and attributes) of ethnic kin across the
border. They show that the presence of ethnic kin across the border
has a strongly significant and large, curvilinear effect on the probability of conflict. The risk of conflict increases as the number of ethnic kin
increases but only to a certain point, after which it declines – the result,
they argue, of the incumbent’s strategically negotiating (and thereby
avoiding conflict) with groups that have powerful kin across the
border. The transborder effect is limited to groups that are excluded
in the neighboring state. Stateless groups such as the Kurds are more
likely to launch armed conflict: “Thus, the pernicious effect of political
exclusion has a tendency to spill over state borders” (26).
While these findings clarify the conditions under which transborder
ethnic groups contribute to conflict, the precise nature of the contribution of co-ethnics in the other state is not clear. How precisely does
the presence of co-ethnics contribute to conflict but only when the
group is a large, excluded minority? Do the neighbors support
conflict but only when they weigh the prospective costs as outweighed
by the benefits given the size of their excluded co-ethnics? Are the
neighbors mobilized into such support by rebel agents? If so, is their
decision better captured as emulation or bounded learning, rather than
rational learning? Might they be coerced by threats against family
across the border?
Sometimes such transborder ties are constructed by conflict, or the
threat of conflict, as when civilians flee conflict in one state across the
borders of another. The presence of refugees may contribute to conflict
in the host state for various reasons, according to Salehyan and
Gleditsch (2006). With refugees may come rebels, arms, and supply
networks; in addition, a large presence of refugees may threaten the
ethnic balance in the host state, or impose externalities on the host, for
example, as when competition for scarce resources increases. In their
statistical analysis, they confirmed that the presence of refugees led
to a significant and substantial increase in risk of conflict in the host
country (controlling for both domestic risk factors in the host and the
242
Elisabeth Jean Wood
presence of conflict in neighboring states). However, other studies
have failed to confirm this finding. Buhaug and Gleditsch (2008) found
that refugees have no effect on the risk of conflict in the host in general.
When they disaggregated the dependent variable into conflicts over
territory versus conflicts over control of the center, they found that
the presence of refugees did affect both, but in opposite directions.
Focusing on the effect of refugees on conflict in their state of origin
(not the host), Salehyan (2007) found that their presence affected the
duration but not the onset of conflict, and did so only when they were
present in unstable states hostile to their state of origin.
Thus, the causal mechanism linking the presence of refugees to
increased conflict remains insufficiently articulated in the quantitative
literature. A distinct mechanism is the presence of rebel bases across the
border. Salehyan (2007) found that it strongly increased the duration
of conflict in the state of origin. Moreover, Salehyan (2008) found
that the presence of rebel bases in one country strongly increases the
risk of inter-state conflict between the host state and the rebels’ state of
origin, particularly when the two states are long-standing rivals.
Another configuration linking state actors to conflict actors is the
support for insurgents by an external state. In their dyadic analysis of
which rebel groups get such external support, Salehyan, Gleditsch, and
Cunningham (2011) found that groups were more likely to receive
military or financial support from the host state if either there was a
transnational constituency or audience (a social group sympathetic on
either ethnic or ideological grounds with the rebels) in the host state, or
the host state was a long-standing rival of the conflict state. The effect
was stronger in the case of territorial conflicts and when the non-host
state itself receives external assistance (presumably from a third state).
Such support tends not to go, however, to rebel groups that are either
quite strong or weak compared to the enemy state.
As may be evident, most of these studies assess the correlation
between conflict diffusion and the conditions hypothesized to contribute
to that diffusion, rather than the process of diffusion itself. To increase
the support for the claim that the hypothesized mechanism indeed
underlies the empirical finding, scholars sometimes complement the
quantitative analysis with qualitative evidence. For example, in an
appendix, Salehyan, Gleditsch, and Cunningham (2011) report the
results of their analysis of case narratives and confirm that, in 80.3
percent of the cases, the expected states (those with transnational
Where do we go from here?
243
constituencies or rivals) were in fact the ones providing rebel support.
To address endogeneity issues, some authors explicitly test the reverse
hypotheses – for example, in Salehyan’s study of rebel bases (2008),
that foreign ties predict rebel strength, rather than the reverse – to rule
out inverse causality, or more minimally lag the left-hand side variables.
Despite these significant advances in the analysis of the transnational
diffusion of conflict, the precise causal mechanisms underlying diffusion
are not yet adequately clear. The findings often identify the conditions
under which diffusion is likely (the presence of transborder ethnic ties,
for example), without demonstrating precisely what mechanism drives
the effect across borders. While new data sets of theoretically relevant
variables mark a significant advance over earlier proxies, this literature
remains largely driven by structural configurations and deploys betterspecified but as-yet inadequate proxies, such as demographic size as a
measure of mobilizational capacity. The process of conflict diffusion
is difficult to observe much less capture in a cross-national database.
The challenge is to identify not only the relevant conditions but also to
specify how particular configurations of actors affect the dynamics of
conflict. This is where the chapters of this volume make their most
important contribution, to which we now turn.
The contributions of this volume
This volume contributes to our understanding of the transnational
dimension of conflict through a focus on causal mechanisms, the
specification of their observable implications, and a careful tracing
of the evidence for those mechanisms in the available data. In short,
the chapters focus on how transnational factors have causal effect. The
challenges to doing so are compounded by the covert nature of many
of the transnational networks and flows, the difficulty in accessing
key insiders, and in many cases the fact that the views of powerful
actors dominate existing narrative accounts of conflicts. In light of
these challenges, essential to the persuasiveness of this type of research
is the identification of rival mechanisms and their observable implications, and tracing the competing evidence for those mechanisms. In the
following assessment, I first highlight the authors’ contributions to our
understanding of the transnational dynamics of civil conflicts, focusing
on transnational mobilization and transnational insurgents, then turn
to the lessons for the analysis of diffusion and qualitative methods.
244
Elisabeth Jean Wood
The transnational mobilization of support
In her chapter on diasporas and civil conflict, Adamson identifies
three mechanisms whereby diasporas may be mobilized in support
of armed actors in their homeland, and two whereby that mobilization shapes a conflict’s course. The first of the three mobilization
mechanisms is transnational brokerage: conflict entrepreneurs link
diaspora networks with conflict actor networks. In the case of Kurdish migrants from Turkey in Germany, some entrepreneurs were
migrants who drew on their personal ties to conflict actors. Others
were PKK activists who traveled between Turkey or Syria and Europe
for the purpose of building support for the Kurdish armed struggle.
These brokers engaged in a variety of activities to further the cause of
Kurdish separatism.
These conflict entrepreneurs engaged in strategic framing, the second
mechanism, promoting Kurdish language and culture to strengthen
Kurdish over Turkish (or German) identity using transnational frames
such as kinship and duty. The framing was effective, she argues,
because many migrants were doubly marginalized: as Turkish immigrants they were denied German citizenship; as Kurds they were also
marginalized in Turkey. As a result, the frame of resistance resonated
in the diaspora, a core requirement for effective transnational mobilization. The PKK came to dominate Kurdish politics in Europe through
a process of ethnic outbidding, the third mechanism. As a result of their
double marginalization, the radical politics of the PKK came to be seen
as more legitimate than the more moderate approaches of their rivals.
The mobilization was very effective in both generating significant
resources for the PKK and in advocating for Kurdish interests in Europe,
the two mechanisms – lobbying-persuasion and resource mobilization –
whereby Adamson argues that diaspora mobilization can affect the
conflict. On the former, PKK activists in the diaspora built civil society
organizations to lobby European governments through both informal
civic activities, such as cultural festivals, and formal institutional politics,
such as developing a government in exile to facilitate the lobbying of
politicians and building ties to leftist parties. Such initiatives often
focused on Turkey’s human rights violations. Regarding resource
mobilization, the PKK developed covert networks to collect
donations, extort “taxes,” and recruit militants, getting these
resources to Damascus and into Turkey through a variety of channels.
Where do we go from here?
245
This analysis illuminates the mechanisms underlying the findings in
the quantitative literature about the role of transnational ethnic ties.
Kurds are an excluded, sizeable minority in Turkey, and so confirm the
results of Cederman, Girardin, and Gleditsch (2009) discussed above.
Such ties did not play a role in and of themselves, however; essential
was a process of active mobilization by political entrepreneurs who set
out to do just that – despite the lack of a common border. Mobilization
along militant lines was thus the result of political agency by entrepreneurs, not something given by the fact of ethnic ties. Moreover, the
recognition of such ties was itself a result of political agency, as many
migrants from southeast Turkey who thought of themselves as Turks
discovered their “Kurdishness” as a result of and through the process
of mobilization. Thus Adamson’s analysis shows how those ties
played a role in both the mobilization of the diaspora and its effect
back in Turkey. The mobilization follows the “boomerang” pattern
of transnational activism laid out by Keck and Sikkink (1998), but
with the difference that it originates with an insurgent rather than a
civic organization.The case also raises further questions. The success
of mobilization in Europe appears to depend critically on democratic
practice there. Activists developed civil society organizations and also
covert networks that engaged in a very wide range of activities, taking
full advantage of associational freedoms and often appealing to human
rights and self-governance global norms. Moreover, the rhetoric of
campaigns often played on the complicated geopolitics between
Europe and Turkey, a NATO member aspiring to European Union
membership. The patterns of mobilization of the Salvadoran diaspora
in the United States in support of Salvadoran insurgents follows
strikingly similar patterns (Perla 2008), with a similar mix of overt
and covert tactics and the leverage of human rights abuses by a state
allied with the US. To what extent does successful mobilization in the
diaspora depend on such democratic space? On the geopolitical issues
of the day?
In his analysis of the demobilization of child soldiers by the separatist movement of southern Sudan – the Sudan People’s Liberation
Army (SPLA) – Hamberg also focuses on the process of transnational
mobilization, but in a case with a distinct configuration of actors. In
this instance, the transnational ties were originally between Christian
congregations in southern Sudan and evangelical Christian organizations in the US. In light of increasing human rights abuses of civilians
246
Elisabeth Jean Wood
by the SPLA, the New Sudan Council of Churches both strengthened
transnational ties to co-religionists in order to indirectly pressure the
insurgent group from the US and also pressed the SPLA directly. In a
context of increasing support from not only US civil society but also
nationally prominent politicians, the SPLA announced that it would
demobilize its child soldiers, and proceeded to do so over the ensuing
months and years.
The insurgents were not responding to “naming and shaming,” as in
the usual model of transnational advocacy, Hamberg argues. He
points out that naming and shaming campaigns were not particularly
targeted at the SPLA and that, when they did mention Sudan, were
focused on the government. Such campaigns thus do not explain why
the SPLA was the only active rebel group to demobilize child soldiers.
The SPLA, he suggests, responded to promises of concrete support, a
“modified boomerang” model.
This case raises several questions about transnational mobilization
by co-religionists in the context of civil conflict. While other cases of
such mobilization have certainly occurred – perhaps most well documented in the mobilization of progressive religious congregations in
the US by their counterparts in Central America during the 1980s
(Smith 1996) – such campaigns have generally targeted governments
rather than rebel groups. Under what conditions do religious networks
attempt to mobilize transnationally to affect the behavior of rebel
groups? And under what conditions does such mobilization succeed?
Hamberg himself emphasizes the domestic attempt by the Council
of Churches to directly influence the SPLA (thereby modifying the
boomerang), which may suggest that this particular pattern may be
most likely in the case of separatist conflicts based on a religious
cleavage. But the analysis leaves other questions unanswered. Why
was the culmination of the campaign the demobilization of child
soldiers rather than some other change of SPLA policy? To what extent
was demobilization a core demand of the Council, and to what extent
was it a positive side effect of a broader campaign for negotiations
to end the war? Moreover, to what extent was the learning by US
Christians about the situation in southern Sudan shaped by their south
Sudanese counterparts, and therefore likely to have been boundedly
rational? How were the ties between the groups established in the
first place? Were the US groups driven only by sympathy for their
co-religionists in Sudan or did the opportunity to influence US policy
Where do we go from here?
247
more broadly also play a role? How were the transnational religious
contacts initiated? What precisely were the relations between the south
Sudanese congregations and the SPLA, and how did they change as a
result of the former’s transnational ties?
Transnational insurgents
In her analysis of the role of transnational insurgents in Chechnya,
Bakke focuses on domestic, rather than transnational, mobilization.
After documenting the role of Arab and other transnational insurgents
in the founding of training camps, mosques, and schools, and sometimes advising key Chechen commanders, Bakke analyzes their effect on
the mobilization of Chechens. She identifies three possible mechanisms:
Transnational insurgents might cause shifts in framing, tactics, or
resource mobilization. She argues that the increasingly Islamic (as
opposed to nationalist) framing of the Chechen rebellion is best
explained by the influence of the transnational insurgents. She is more
skeptical about their influence on the radicalization of tactics by the
rebels (attacks on civilian targets outside of Chechnya, kidnaping, and
suicide attacks), but on balance suggests that transnational insurgents
likely played a role in this area as well, as there was no precedent for
suicide attacks in Chechen tradition. Transnational flows of financial
resources as well as recruits clearly contributed to the funding of
training camps, schools, and mosques (and thereby contributed to
shifts in framing and tactics).
Bakke carefully traces the limits of the role of transnational insurgents as well as their contribution. At least initially, their influence was
dependent on domestic “gatekeepers,” whose permission was necessary
for their presence. Moreover, she traces the domestic roots of some of
the radical tactics such as kidnaping and attacks on civilians to raise the
cost of occupation, showing that only in the adoption of suicide attacks
was the influence of the transnationals indisputable. Finally, given
that the significant financial flows laid out incentives for adopting an
Islamist framing, she acknowledges that the shift in framing may have
been strategic emulation rather than genuine learning.
Bakke’s analysis illuminates the mechanisms whereby such transnational insurgents may affect conflict dynamics. The ability to channel
resources, both material and ideational, appears to have been essential
to the growing ties between the transnational insurgents and Chechen
248
Elisabeth Jean Wood
rebels. Yet ideational resources were controversial within the Chechen
movement. Indeed, the later split in that movement indicates their
resonance with domestic culture was only partial. The influence on
Chechen mobilization was clearest in the shift in framing (if genuine
and not strategic emulation) towards an Islamist justification, raising
once again the role of broadly similar religious culture in the success
of transnational mobilization in this case as well as in southern Sudan.
Would they have been as effective in shifting tactics if from a different
global religion? If transnational mobilization had been limited to
ideational resources, would the transnational insurgents have been
as effective? To what extent was the shift in framing from Chechen
nationalism to Islamist jihad contested within insurgent organizations?
Harpviken and Lischer analyze a distinct pattern of transnational
movement, focusing on refugee militants who engage in violence
on return to their home countries. With both the return of Rwandan
refugees from Uganda in 1990–1994 and Afghan refugees from
Pakistan in 1992, they analyze a “transnational complex,” or a history
of interaction before the return between the refugees and various
actors in their host countries. They discern three central mechanisms
that lead to refugee militancy on return. The first is socialization. In
both cases, many refugees were deeply socialized toward militancy
during exile as they were often socially isolated in conditions of great
uncertainty. The internalization and habituation to militancy in each
case included battlefield socialization (fighting with Ugandan rebels for
many Rwandan refugees and in trips to fight in the homeland for
Afghans in Pakistan). The second is resource control. In Pakistan,
particular political organizations (or their armed wings) controlled
resources and founded both secular and religious schools as well as
military training camps. On return, the urgent need to access resources
to rebuild livelihoods reinforced ties to such organizations, which were
often the only access to land or to state resources in the fragile and
uncertain post-conflict conditions. The third is security entrapment.
On return, many refugees faced escalating security threats as competition for access to resources increased, with the result that many sought
protection with a familiar militarized group.
These mechanisms also account for local support for the resurgent
Taliban after the US invasion in 2001. Most returned refugees did
not support the Taliban after the US invasion, but as collateral damage
from US attacks accumulated (combined with the inability of the
Where do we go from here?
249
Afghan government to deliver on anything from basic services to
security), they rejoined the Taliban as the best way to address their
security dilemma. (The authors also show that Hutu extremists, after
being forced from Rwanda into the eastern Democratic Republic of
Congo [DRC], attempted to build institutions to socialize the mainly
Hutu refugees through their control of resources, but those efforts
were cut short when Rwanda invaded and forced many to return
to Rwanda, where they were put in camps to re-socialize them.)
Harpviken and Lischer thus identify a distinct mechanism underlying
the connection between refugees and conflict – one not discussed in any
detail in the quantitative literature – namely, the return of militant
refugees. Their analysis suggests that returnee violence – the dark
side of refugee agency – is most likely when refugees are socialized
by militant groups in exile, particularly if it includes battlefield
socialization, and when they rely on armed groups for protection or
access to resources on return. However, the analysis does not explore
the conditions under which the return will be as a coherent, effective
militarized organization, as in Rwanda, rather than a disparate set of
factions, as in Afghanistan in 1992 (but not after 2001). Whether or
not the three mechanisms account for this divergence is not clear.
Schmitz analyzes the case where a rebel group itself became transnational. The Lord’s Resistance Army is a hard case for showing
transnational influences, for the reasons he points out but also because
the group appears to have engaged in little mobilization of the diaspora
or built an overt transnational network of any kind. On the other
hand, the group itself has been based in a series of distinct countries,
moving from its origins in Uganda to southern Sudan, then to northeastern DRC via Uganda, and then to the Central African Republic.
Schmitz argues that even in this case, transnational mechanisms had
an important but not necessarily overwhelming role in shaping the
course of the conflict.
Like Bakke, Schmitz analyzes transnational influences in terms of
shifts in framing, tactics, and resources. He traces the origins of the
group’s ideology to the transnational influence of Christian missions,
lending a messianic cadence to its beliefs. While in southern Sudan,
where the group engaged in operations with Sudanese government
forces (and pro-government militias), it widened its repertoire to
include mutilation of supposed collaborators. In response to increasing
criticism from human rights groups and the Acholi diaspora, the group
250
Elisabeth Jean Wood
shifted both its repertoire (with mutilation nearly disappearing) and
framing (shifting away from religious to more secular, human rights
focused rhetoric). Resource mobilization had long been transnational
for the LRA, with access to bases and financial resources from Khartoum
as part of its policy of arming the group as a proxy force against the
SPLA. Despite these transnational links, Schmitz is careful to point
out that the influence of diaspora politics is much less than in the Kurdish
case analyzed by Adamson.
Schmitz analyzes the international prosecution of armed group
leaders as a distinct causal mechanism shaping transnational conflict
dynamics. In his analysis of the failure of peace negotiations, Schmitz
emphasizes the negative effects of the indictment of LRA leaders by the
International Criminal Court (originally, on the government of
Uganda’s invitation). He interprets international prosecution as a
form of non-relational diffusion, arguing that – in the absence of the
personal ties of relational diffusion – it may create sharp disincentives
for conflict termination.
The case appears to be a compelling instance of the arming of an
insurgency (and allowing base camps) by a rival state, in this case
Sudan. But as is perhaps inevitable in this hard-to-document case,
Schmitz’ analysis begs more questions than the available evidence
can address. Was Sudan’s target the government of a rival state, or
its own insurgent group? The evidence appears to support the latter
given the record of activity in Sudan that Schmitz presents. Under what
conditions will a state arm the insurgents of a rival state as a proxy
against its own insurgents? Was the widening of the LRA’s repertoire
after going on operations with Sudanese forces an instance of
emulation, as he argues, of bounded learning, or perhaps of coercive
learning in response to conditioning of aid from Sudan in some way?
Emulation or bounded learning appears equally plausible given
the limited evidence. Under what perhaps implicit conditions did the
government of Sudan arm the LRA, and what precisely undermined
that relationship?
On transnational diffusion
The chapters thus contribute to our understanding of the causal
mechanisms underlying transnational diffusion by showing how
the dynamics of conflict in one country are shaped by dynamics in
Where do we go from here?
251
another. The mechanisms include most of those analyzed in the policy
diffusion literature (emulation, learning, and shifting incentives) and
in the quantitative literature (the presence of transnational ethnic ties,
refugees, and rebel bases) and thus add to our understanding of the
“dark side” of transnational politics under which it is not positive
global norms but conflict dynamics that shape the behavior of conflict
actors. Moreover, the chapters also identify new mechanisms of
transnational diffusion and demonstrate their importance for understanding conflict dynamics.
More specifically, one theme of the volume is the extent to which
migration is an essential mechanism of diffusion, a mechanism little
emphasized in the policy literature. In Chechnya, transnational insurgents brought a variety of material and ideational resources to separatist
rebels. In Europe, PKK members founded civil society organizations as
well as covert networks. In both Rwanda and Afghanistan in the early
1990s, militant refugees returning from exile engaged in battles for
control of their homeland state as well as for control of territory and
resources. In Sudan, the DRC, and the Central African Republic as well
as Uganda, the LRA carried out its repertoire of violence, particularly
looting and abductions. The LRA case also points to the importance of
a multiplicity of state borders for an insurgent group seeking refuge.
A second, broad mechanism is the mobilization of external populations. As Adamson demonstrates in her chapter on the Kurdish/Turkish
civil conflict – and in contrast to the quantitative literature which
generally treats diasporas as unitary actors predisposed to become
active in conflicts – diasporas must be actively mobilized through the
usual contentious politics mechanisms of framing, tactical innovation,
and resource mobilization. Hamberg’s chapter shows that non-diasporic,
external groups may be mobilized as well, particularly religious
networks in support of actors fighting against discrimination and violence targeted at members of that religion. The chapter demonstrates
the importance of personal contact for such mobilization, between representatives of the two religious networks but also between the politicians
lobbied by each one. Mobilization in this case led to a lessening of conflict
both in the form of child soldier demobilization and arguably in the
form of increased international support for a negotiated resolution of
the decades-long conflict.
Similarly, if refugee populations are to actively shape the conflict in
their homeland, they must be mobilized, as Harpviken and Lischer show.
252
Elisabeth Jean Wood
The conditions are, however, fairly stringent, according to their
analysis. Militant groups must control sufficient resources in exile
(to monopolize institutions through which they can effectively socialize
the refugee population) and also on return (to enforce group loyalty
through the control of scarce resources), and the returnees must face a
security trap (and thus redouble their allegiance to the militant group).
Nome and Weidmann’s agent-based model confirms the efficacy of
brokerage by transnational norm entrepreneurs as a mechanism for
mobilizing transnational populations. The model focuses on the role
of such brokers in laying the groundwork for mobilization through
changes in the repertoire of imagined social identities. In circumstances
where signals from the homeland are noisy (likely the case from a country
in conflict), brokerage is significantly more effective in diffusing the
new, though latent, social identify. That identity could then become
active (and thus mobilizeable) under conditions of not unreasonably
high receptivity among the diaspora. The cases of the actively mobilized
Kurdish and minimally mobilized Acholi diaspora (the ethnic group
on which the LRA originally drew) appear to be positive and negative
instances of this mechanism.
The model also raises the question of mobilization of a diaspora in
support of the state rather than the insurgency, as nothing in the model
presumes the brokers are rebels. Hector Perla Jr., (2005; 2009) demonstrated just such a pattern – which he terms the “signal flare”
mechanism – in the case of Sandinista Nicaragua mobilizing Nicaraguans in the US to oppose US policy in Central America, particularly
US support for the contra insurgents.
The chapters thus contribute to research on transnational contentious
politics through the identification and demonstration of the causal
force (under certain conditions) of the mechanisms highlighted in that
literature. Tarrow (2005) identifies three classes of transnational diffusion mechanisms: relational (diffusion through personal ties), mediated
(through the brokering of a mutual contact), and non-relational (through
impersonal means such as learning from publications or news reports).
Bakke emphasizes the importance of relational diffusion (personal
ties between insurgent leaders and a Chechen commander) early in her
Chechen case. Transnational brokers, as Adamson demonstrates, were
essential to the mobilization of the Kurdish diaspora in Europe.
The role of non-relational diffusion through global media of innovations in framing, tactics, and resource mobilization, which may occur
Where do we go from here?
253
alongside relational and brokered diffusion, is much more difficult to
assess. To what extent did jihadist videos available via DVDs and the
internet (specifically, those not distributed by the schools, mosques,
and camps founded by the transnational insurgents themselves) also
influence the shift toward Islamist framing from that of Chechen
nationalism in Bakke’s study? The chapters thus deepen our understanding of transnational diffusion by disaggregating that overly broad
concept into several analytically distinct mechanisms – and sometimes
disaggregating those, as when mobilization of a diaspora is understood
as a composite mechanism made up of the contentious politics
mechanisms of framing, tactical innovation, and resource mobilization.
As noted by Checkel (this volume), the proper level of disaggregation is
given by the state of the literature. One measure of this volume’s advances
in the understanding of diffusion is the disaggregation of mechanisms
along with the (often implicit) statement of when each is likely.
On method
To address the challenging task of identifying causal mechanisms that
shape transnational conflict dynamics, the chapters engage in process
tracing (with the exception of the chapter by Nome and Weidmann
to which I return below). Process tracing involves a commitment to
showing how a result was brought about. To persuasively show that a
particular mechanism was in fact the “how” that brought about
the result in the difficult setting of transnational conflict dynamics,
an author must identify a plausible mechanism and its observable
implications; tell a compelling, well-evidenced narrative documenting
that those observable implications in fact occurred (subject to the limitations and bias of the relevant data), and also show that the observable
implications of rival mechanisms do not occur after “relentlessly”
(Bennett, this volume) pursuing that data.
In working with incomplete, indeed sometimes fragmentary, data,
the assessment of the likely bias in the source is, of course, crucial.
Moreover, arguments based on process tracing should account for
important turning points in the sequence of events (Bennett, this
volume). Given the limitations of data, it may emerge that while no
“smoking gun” was found, the explanation based on that mechanism
gains plausibility if alternatives are not supported (or are significantly
less well supported).
254
Elisabeth Jean Wood
Hamberg’s chapter stands out for his careful construction of a
difficult argument despite the lack of a smoking gun, which in his case
would have been direct confirmation that in fact the SPLA demobilized
child soldiers in exchange for support from US political actors. No
such datum was forthcoming, but Hamberg persuasively shows that
arguments based on rival mechanisms – the SPLA no longer needed
child soldiers, that NGO or UN naming and shaming campaigns
calling for demobilization of child soldiers had made their retention
untenable, among others – are unpersuasive. Bakke also carefully
weighs the evidence for the domestic origins of the shifts in framing,
tactics, and resources she documented, reaching a nuanced judgment
that the influence of transnational insurgents varied across those three
issues. Schmitz discusses the authenticity of manifestos supposedly put
out by the LRA, concluding that, while it was not possible to attribute
authorship to particular group leaders, the documents nonetheless
reflected the organization’s views at the time. Harpviken and Lischer
take advantage of the additional analytic leverage available through
the comparison of two cases of refugees who engaged in violence
on their return. The divergent ways refugees were socialized in exile
(Afghans largely residing in camps controlled by militant factions
and immersed in institutions for military and political socialization;
Rwandans living in isolation without such immediate dominance
of everyday lives), allowed the authors to focus on particular aspects
of exile socialization, and argue for the causal centrality of these
common factors. For example, the role of battlefield socialization is
persuasive in part because it was present in both cases. The authors
also leveraged internal comparison to some degree, comparing those
Afghan factions who fought for Kabul in 1992 with those who did
not, the return of Afghan refugees in 1992 and after 2001, and the
return of Rwandan Tutsi refugees from 1990 to 1994 with the coerced
return of Rwandan Hutu refugees after 1996.
Nome and Weidmann in their chapter demonstrate how agent-based
modeling complements both process tracing and quantitative analysis.
After noting that quantitative studies usually only capture the conditions that are said to result in diffusion or the fact of outcome (but
not the process), they model two mechanisms whereby transnational
diffusion of social identity might occur. They are able to show through
their analysis of simulations that one mechanism (brokerage in the
form of a transnational norm entrepreneur, followed by social
Where do we go from here?
255
adaptation) more consistently results in successful diffusion than
the other (adaptation across country borders). They are careful to
label the result a confirmation of the one mechanism as a “candidate
explanation” subject to further empirical testing.
The chapter shows how the method forces its practitioners to clearly
lay out the logic of the mechanism in the form of the rules that the
agents follow (with some degree of error) and nonetheless produces
results that are not obvious beforehand. In particular, drawing on
the logic of the brokerage-followed-by-adaptation mechanism, Nome
and Weidmann show that a change of social identity on the part of
both young Chechens and young Kurds would have been likely, not
simply the shift in framing traced by Bakke in Chechnya nor the
promotion of new cultural tropes by PKK activists in Europe, as
discussed by Adamson.
Process tracing as conducted in this volume thus demonstrates
the power of the method when deployed with careful assessments of
rival mechanisms. The chapters also demonstrate how the method is
complementary to quantitative analysis (Sambanis 2004; George and
Bennett 2005; Gerring 2007b) and agent-based modeling (Nome and
Weidmann, this volume), confirming some mechanisms discussed in
the quantitative literature, identifying new ones, and fleshing out
the actors, motives, alliances, and the intended and unintended consequences of their actions.
Conclusion: a research agenda
How might the above contributions to our understanding of transnational conflict dynamics, transnational diffusion, and qualitative
method be further leveraged to advance that understanding still more
deeply? Moving from the identification of causal mechanisms towards a
theory of transnational conflict dynamics depends on the specification
of scope conditions in the form of hypotheses and then the aggregation
of distinct hypotheses into a coherent theory. Such a task is far beyond
what this chapter can achieve. In what follows, I take up a more feasible
task, suggesting aspects of the research agenda implied by this volume’s
essays. Just as scholars of policy diffusion are beginning to study how
learning is conditional on the politics of both source and innovating
country – the who-learns-what-from-whom approach mentioned above –
one promising way forward for students of transnational conflict
256
Elisabeth Jean Wood
dynamics is to interrogate more carefully the conditions under which
diasporas are mobilized, coalitions constructed, and resources dispersed.
We saw above that the successful mobilization of the Kurdish
diaspora in Europe was in part due to the democratic nature of
European governance (particularly freedom of association). Moreover,
that mobilization leveraged (more or less successfully at different times)
both the global normative discourse of human rights and also the
geopolitical ties between Europe and Turkey. This suggests the
following hypothesis:
The more democratic is the host state and the more allied it is with the origin
state, the more likely is mobilization of support in the diaspora by insurgents.
It also begs further questions. Under what conditions will the state
itself attempt to mobilize support in the diaspora? And when do
contending efforts to mobilize the diaspora arise, with one group of
transnational entrepreneurs attempting to mobilize the diaspora in
favor of one group (perhaps the insurgents) and another in favor of
another group (a rival non-state group, or perhaps the state itself)?
As discussed in the chapter by Adamson, the PKK emerged as
the dominant force in the Kurdish diaspora because its more radical
message resonated more deeply with the doubly marginalized Kurds in
Germany than that of its moderate rivals. This suggests the following
hypothesis:
The more democratic is the host state and the marginalized are members of
the diaspora in the state of origin, the more likely is mobilization of the
diaspora (perhaps by contending actors).
However, the mobilization of Nicaraguan diaspora in the US by the
Sandinista government (Perla 2005; 2008) suggests that such mobilization might also be likely when the states are themselves in conflict,
leading to the more complex hypothesis:
The more democratic is the host state and the more conflictual are the host
and origin states, the more likely mobilization of the diaspora will occur by
the state of orgin.
Another question raised by the chapters on southern Sudan and
Chechnya is the role of a common religion in the building of transnational
alliances. Would the transnational insurgents have been as welcome
in Chechnya if Islam had not been a shared religion (notwithstanding
Where do we go from here?
257
the many differences between the Salafi approach to Islamist thought
and that of Chechen Muslims)? Similarly, the alliance between southern
Sudanese Christian churches and evangelical networks in the US appears
to have depended to a large degree on a shared Christian faith, framed as
under attack in southern Sudan. Are conflicts along the lines of a religious
cleavage between two world religions particularly likely to draw in
transnational allies, either for mitigating conflict as in this case, or for
stoking it, as in the former? This suggests the following hypothesis:
Transnational mobilization of allies is more likely in the case of conflicts
along the lines of a religious (or ethnic) cleavage than where that cleavage
is absent.
Note that both these conflicts are separatist conflicts, so one might
argue more specifically that: transnational mobilization of allies is
more likely in the case of separatist conflicts along the lines of a
religious cleavage.
The volume’s emphasis on migration as a mechanism of transnational diffusion (one nearly absent in the literature on policy diffusion) similarly suggests research questions and hypotheses. Are the
mechanisms identified in the Rwandan and Afghan cases of refugee
militants relevant for other instances of refugee militancy? This leads to
the following hypothesis:
Where militant groups control schools in exile and promote military
training, refugees themselves participate in military excursions, and on
return such groups control access to livelihood and security, refugees are
more likely to participate in violence.
Raised but not fully answered in the Rwandan/Afghan comparison
is an additional research question: given the conditions for refugee
militancy, what are the further conditions for the emergence of a
coherent, unitary organization in exile (versus a variety of factions)?
The agent-based model of transnational diffusion of social identity
suggested that transnational norm entrepreneurs are more efficient in
such diffusion than mere social adaptation given that the “signal”
coming from a conflict-ridden country is likely complex (noisy) and
assuming that the receptivity on the part of the diaspora population
to the new identity is not too low. As a candidate explanation, the
implications call for further research on cases where mobilization
of a diaspora occurred – and also where it did not despite efforts of
258
Elisabeth Jean Wood
either type. More generally, the chapter highlights the usefulness of
agent-based modeling in clarifying the logic of causal mechanisms
through precisely stipulating this logic and simulating the consequences.
The comparison of two candidate mechanisms in the same modeling
framework in the chapter is particularly useful.
The analysis of the diffusion of violent tactics – for example, suicide
terrorism into Chechnya – suggests that a broader study of the diffusion
of distinct patterns of insurgent violence would be a timely research
initiative, complementing the many studies on terrorism. Ideally, an
investigation would document and analyze the conditions for the transnational diffusion of targeting patterns and of forms of violence
(massacres as retaliation, disappearances, particular forms of torture,
etc.) as well as the underlying mechanisms, including migration of
combatants, emulation, learning from peer organizations or from media,
coercion, and competition. Although it is more difficult to gather data on
non-lethal than lethal injuries, scholars and activists increasingly collect
data on torture, sexual violence, abduction, and other types of violence
(Hafner-Burton and Ron 2009). Additionally, disaggregated data about
the pattern of attacks is increasingly available, as in the recently released
Uppsala Conflict Data Program – Georeferenced Event Dataset, which
covers Africa (see Sundberg, Lindgren, and Padskocimaite 2010).
While the chapters largely focus on transnational dynamics – that is,
relations across borders between non-state actors and between non-state
actors and states – the empirical reality of transnational mechanisms
linking insurgencies and external states to another state’s insurgency
raises the issue of the contribution of state-to-state relations to transnational dynamics of conflict. It is likely that an adequate understanding
of transnational dynamics will be impossible without an understanding
of state-to-state innovation, many times in response to transnational
developments – a crucial analytic move increasingly seen in the broader
transnational politics literature as well (Orenstein and Schmitz 2006).
More broadly still, further development and testing of the mechanisms
underlying transnational conflict dynamics will likely come from analysis of a broader range of cases, particularly those from the Cold War
era, including national liberation movements, revolutionary armed
struggles, and the transnational dynamics of states opposing both.
Bibliography
Abdelal, R., Y. M. Herrera, A. I. Johnston, & R. McDermott, 2009. “Identity as a Variable.” In R. Abdelal, Y. M. Herrera, A. I. Johnston, &
R. McDermott (eds.), Measuring Identity: A Guide for Social Scientists.
Cambridge University Press.
Acharya, Amitaev. 2004. “How Ideas Spread: Whose Norms Matter? Norm
Localization and Institutional Change in Asian Regionalism.” International Organization. 58 (2), 239–275.
Achverina, Vera & Simon Reich. 2009. “No Place to Hide: Refugees,
Displaced Persons, and Child Soldier Recruits.” In Scott Gates & Simon
Reich (eds.), Child Soldiers in the Age of Fractured States. Pittsburg
University Press.
Adam, Jeroen, Bruno De Cordier, Kristof Titeca, & Koen Vlassenroot. 2007.
“In the Name of the Father? Christian Militantism in Tripura, Northern
Uganda, and Ambon.” Studies in Conflict & Terrorism. 30, 963–983.
Adamson, Fiona B. 2002. “Mobilizing for the Transformation of Home:
Politicized Identities and Transnational Practices.” In Nadje al-Ali &
Khalid Koser (eds.), New Approaches to Migration? Transnational
Communities and the Transformation of Home. London: Routledge,
155–168.
2004. “Displacement, Diaspora Mobilization, and Transnational Cycles of
Political Violence.” In John Tirman (ed.), Maze of Fear: Migration and
Security after 9/11. New York: New Press, 45–58.
2005. “Globalization, Transnational Political Mobilization, and Networks of Violence.” Cambridge Review of International Affairs.
18 (1) April, 35–53.
2006. “Crossing Borders: International Migration and National Security.”
International Security. 31 (1) Summer, 165–199.
2012. “Constructing the Diaspora: Diaspora Identity Politics and Transnational Social Movements.” In Terrence Lyons & Peter Mandaville
(eds.), Globalization and Diaspora: Politics from Afar. London: Hurst
and Co., 25–42.
Adamson, Fiona B. & Madeleine Demetriou. 2007. “Rethinking the Boundaries Between ‘State’ and ‘National Identity:’ Incorporating Diasporas
259
260
Bibliography
into IR Theorizing.” European Journal of International Relations.
13 (4) December, 489–526.
Adelman, Howard. 2002. “Refugee Repatriation.” In Stephen John
Stedman, Donald Rotchild, & Elizabeth M. Cousens (eds.), Ending
Civil Wars: The Implementation of Peace Agreements. Boulder, CO:
Lynne Rienner, 273–302.
Adler, Emanuel. 2002. “Constructivism and International Relations.” In
Walter Carlsnaes, Thomas Risse, & Beth Simmons (eds.), Handbook
of International Relations. London: Sage Publications, 95–118.
Akgündüz, Ahmet. 1993. “Labour Migration from Turkey to Western
Europe (1960–1974): An Analytical Review.” Capital and Class. 51,
153–194.
Allen, Tim. 2005. War and Justice in Northern Uganda: An Assessment of
the International Criminal Court’s Intervention. London: Crisis States
Research Centre/London School of Economics.
2006. Trial Justice. The International Criminal Court and the Lord’s
Resistance Army. London: Zed Books.
Allen, Tim and Mareike Schomerus. 2006. A Hard Homecoming: Lessons
Learned from the Reception Center Process on Effective Interventions
for Former “Abductees” in Northern Uganda. Washington DC/Kampala: USAID/UNICEF.
Allen, Tim and Koen Vlassenroot (eds.). 2010. The Lord’s Resistance Army:
Myth and Reality. London: Zed Books.
Ames, Barry. 2009. “Methodological Problems in the Study of Child Soldiers.” In Scott Gates and Simon Reich (eds.), Child Soldiers in the Age
of Fractured States. Pittsburg University Press.
Amnesty International. 1991. Sudan: Human Rights Violations during the
Military Government’s Second Year in Power. London: Amnesty
International.
1997. “Breaking God’s Commands:” The Destruction of Childhood by
the Lord’s Resistance Army. London: Amnesty International.
1999. Uganda. Breaking the Circle: Protecting Human Rights in the
Northern War Zone. London: Amnesty International.
Amstutz, J. Bruce. 1986. Afghanistan: The First Five Years of Soviet Occupation. Washington DC: National Defense University Press.
Anderson, Benedict. 1998. “Long-Distance Nationalism.” In Benedict
Anderson, The Spectre of Comparisons: Nationalism, Southeast Asia,
and the World. London: Verso, 58–74.
Andvig, Jens Christopher and Scott Gates. 2009. “Recruiting Children for
Armed Conflict.” In Scott Gates & Simon Reich (eds.), Child Soldiers in
the Age of Fractured States. Pittsburg University Press.
Bibliography
261
Angousteres, Aline and Valerie Pascal. 1996. “Diasporas et financement des
conflits.” In F. Jean & J.-C. Ruffin (eds.), Economie des Guerres Civiles.
Paris: Hachette.
Annan, Jeannie, Christopher Blattman, Dyan Mazurana, & Khristopher
Carlson. 2009. “Women and Girls at War: ‘Wives,’ Mothers and
Fighters in the Lord’s Resistance Army.” Mimeo. New Haven: Department of Political Science, Yale University (October).
Arusha Peace Agreement. 1993. “Arusha Peace Agreement: Protocol of
Agreement between the Government of the Republic of Rwanda and
the Rwandese Patriotic Front on the Repatriation of Rwandese Refugees and the Resettlement of Displaced Persons.”
Autesserre, Severine. 2009. “Hobbes and the Congo: Frames, Local Violence
and International Intervention.” International Organization. 63 (2),
249–280.
Baiev, Khassan. 2003. The Oath: A Surgeon Under Fire. London: Simon &
Schuster.
Baltrusaitis, Daniel. 2009. Coalition Politics and the Iraq War: Determinants
of Choice. Boulder, CO: Lynne Reiner.
Barfield, Thomas J. 2010. Afghanistan: A Cultural and Political History.
Princeton University Press.
Barkey, Henri J. & Graham E. Fuller. 1998. Turkey’s Kurdish Question.
Oxford: Rowman and Littlefield.
Barnett, Michael. 1999. “Culture, Strategy and Foreign Policy Change:
Israel’s Road to Oslo.” European Journal of International Relations.
5 (1), 5–36.
Beah, Ishmael. 2007a. “The Making, and Unmaking, of a Child Soldier.”
New York Times. January 14.
2007b. A Long Way Gone: Memoirs of a Boy Soldier. New York: Farrar,
Straus and Giroux.
Behrend, Heike. 1999. Alice Lakwena & the Holy Spirits: War in Northern
Uganda, 1985–97. Oxford: James Curry.
Beissinger, Mark. 2002. Nationalist Mobilization and the Collapse of the
Soviet State. New York: Cambridge University Press.
Belz, Mindy. 2001. “More than a Warlord.” World Magazine. July 21.
Benford, Robert D. 1993. “Frame Disputes within the Nuclear Disarmament
Movement.” Social Forces. 71 (3), 677–701.
Benford, Robert D. & David A. Snow. 2000. “Framing Processes and Social
Movements: An Overview and Assessment.” Annual Review of Sociology. 26, 611–639.
Bennett, Andrew. 2003. “A Lakatosian Reading of Lakatos: What Can we
Salvage from the Hard Core?” In Colin Elman & Miriam Elman (eds.),
262
Bibliography
Progress in International Relations Theory: Appraising the Field. Cambridge, MA: MIT Press, 455–494.
2008. “Process Tracing: A Bayesian Approach.” In Janet BoxSteffensmeier, Henry Brady, & David Collier (eds.), Oxford Handbook
of Political Methodology. Oxford University Press.
Bennett, Andrew & Jeffrey T. Checkel (eds.). Forthcoming. Process Tracing
in the Social Sciences: From Metaphor to Analytic Tool.
Bennett, Andrew, & Colin Elman. 2006. “Qualitative Research: Recent
Developments in Case Study Methods.” Annual Review of Political
Science. 9, 455–476.
2007. “Case Study Methods in the International Relations Subfield.”
Comparative Political Studies. 40 (2), 170–195.
Bennett, Andrew & Alexander George. 2005. Case Studies and Theory
Development in the Social Sciences. Cambridge, MA: MIT Press.
Bennett, Andrew, Joseph Lepgold, & Danny Unger. 1994. “Burden-sharing
in the Persian Gulf War.” International Organization. 48, 39–75.
Bevan, James. 2007. “The Myth of Madness: Cold Rationality and ‘Resource’
Plunder by the Lord’s Resistance Army.” Civil Wars. 9, 343–358.
Bhaskar, Roy. 1975. A Realist Theory of Science. Leeds Books.
Biswas, Bidisha. 2004. “Nationalism by Proxy: A Comparison of Social
Movements Among Diasporic Sikhs and Hindus.” Nationalism and
Ethnic Politics. 10 (2), 269–295.
Black, Richard. 2006. “Return of Refugees: Retrospect and Prospect.” In
Michael Dumper (ed.), Palestinian Refugee Repatriation: Global Perspectives. London: Routledge, 23–40.
Blattman, Christopher. 2007. “The Causes of Child Soldiering: Theory and
Evidence from Northern Uganda.” Paper presented at the Annual Convention of the International Studies Association, Chicago (March).
Blattmann, Christopher & Jeannie Annan. 2010. “On the Nature and
Causes of LRA Abduction: What the Abductees Say.” In Tim Allen &
Koen Vlassenroot (eds.), The Lord’s Resistance Army: Myth and Reality. London: Zed Books, 132–155.
Blattman, Christopher & Edward Miguel. 2010. “Civil war.” Journal of
Economic Literature. 48 (1), 3–57.
Bob, Clifford. 2005. The Marketing of Rebellion: Insurgents, Media and
International Activism. New York: Cambridge University Press.
Bobrovnikov, Vladimir. 2001. “Al-Azhar and Shari’a Courts in TwentiethCentury Caucasus.” Middle Eastern Studies. 37 (4), 1–24.
Borchgrevink, Kaja. 2010. “Religious Education in Afghanistan and
Trans-border Religious Networks.” PRIO Paper. Peace Research Institute
Oslo.
Bibliography
263
Boutroue, Joel. 1998. “Missed Opportunities: The Role of the International
Community in the Return of the Rwandan Refugees from Eastern Zaire:
July 1994–December 1996.” Cambridge, MA: Center for International
Studies, MIT and UNHCR.
Brady, Henry. 2008. “Causation and Explanation in Social Science.” In
Janet Box-Steffensmeier, Henry Brady, & David Collier (eds.), The
Oxford Handbook of Political Methodology. Oxford University Press,
217–270.
Branch, Adam. 2007. “Uganda’s Civil War and the Politics of ICC Intervention.” Ethics & International Affairs. 21, 179–198.
Brinkerhoff, Jennifer. 2006. “Digital Diasporas and Conflict Prevention: The
Case of Somalinet.com.” Review of International Studies. 32 (1),
25–47.
Brown, Elijah, M. 2008. “The Road to Peace: The Role of the Southern
Sudanese Church in Communal Stabilization and National Resolution.”
Ph.D. Dissertation, University of Edinburgh. www.era.lib.ed.ac.uk/
bitstream/1842/3260/1/EM%20Brown%20PhD%20thesis%2008.pdf.
(Accessed July 11, 2010)
Brownback, Sam. 1999. “Brownback Presses for Food Aid to Suffering
Victims in Sudan.” News Release. http://brownback.senate.gov/pressapp/record.cfm?id 175969. (Accessed July 26, 2010)
2007. From Power to Purpose. Nashville: Thomas Nelson.
Brubaker, Rogers. 1996. Nationalism Reframed: Nationhood and the
National Question in the New Europe. Cambridge University Press.
2005. “The ‘Diaspora’ Diaspora.” Ethnic and Racial Studies. 28 (1), 1–19.
Brubaker, Rogers & David D. Laitin. 1998. “Ethnic and Nationalist Violence.” Annual Review of Sociology. 24, 423–452.
Bruce, John. 2007. “Drawing a Line Under the Crisis: Reconciling Returnee
Land Access and Security in Post-conflict Rwanda.” In Humanitarian
Policy Group Working Paper. London: Overseas Development
Institute.
Bruinessen, Martin van. 1998. “Shifting National and Ethnic Identities: The
Kurds in Turkey and the European Diaspora.” Journal of Muslim
Minority Affairs. 18 (1), 39–51.
Buhaug, Halvard & Scott Gates. 2002. “The Geography of Civil War.”
Journal of Peace Research. 39 (4), 417–433.
Buhaug, Halvard & Kristian Skrede Gleditsch. 2008. “Contagion or Confusion? Why Conflicts Cluster in Space.” International Studies Quarterly.
52, 214–233.
Buhaug, Halvard & Jan Ketil Rød. 2006. “Local Determinants of African
Civil Wars, 1970–2001.” Political Geography. 25 (3), 315–335.
264
Bibliography
Bundesamt für Verfassungsschutz. 1996. Die Arbeiterpartei Kurdistans
(PKK) – Strukturen, Ziele, Aktivitaeten. Cologne: Bundesamt für
Verfassungsschutz.
Burt, Ronald S. 2005. Brokerage and Closure: An Introduction to Social
Capital. Oxford University Press.
Busby, Joshua William. 2007. “Bono Made Jesse Helms Cry: Jubilee 2000,
Debt Relief, and Moral Action in International Politics.” International
Studies Quarterly. 51 (2), 247–275.
Byman, Daniel, Peter Chalk, Bruce Hoffman, William Rosenau, & David
Brannan. 2001. Trends in Outside Support for Insurgent Movements.
Santa Monica: Rand.
Capie, David. 2008. “Localization as Resistance: The Contested Diffusion
of Small Arms Norms in Southeast Asia.” Security Dialogue. 39,
637–658.
Caporaso, James. 2009. “Is There a Quantitative-Qualitative Divide in Comparative Politics? The Case of Process Tracing.” In Todd Landman &
Neil Robinson (eds.), The Sage Handbook of Comparative Politics.
London: Sage Publications.
Carpenter, R. Charli. 2007. “Setting the Advocacy Agenda: Theorizing Issue
Emergence and Nonemergence in Transnational Advocacy Networks.”
International Studies Quarterly. 51 (1), 99–120.
Cederman, Lars-Erik. 2001. “Agent-Based Modeling in Political Science.”
The Political Methodologist: Newsletter of the Political Methodology
Section of the American Political Science Association. 10 (1), 16–22.
2003. “Modeling the Size of Wars: From Billiard Balls to Sandpiles.”
American Political Science Review. 97 (1), 135–150.
Cederman, Lars-Erik, Kristian Skrede Gleditsch, Idean Salehyan, & Julian
Wucherpfennig. Forthcoming. “Transborder Ethnic Kin and Civil
War.” International Organization.
Cederman, Lars-Erik, Luc Girardin, & Kristian Skrede Gleditsch. 2009.
“Ethnonationalist Triads: Assessing the Influence of Kin Groups on
Civil War.” World Politics. 61 (3), 403–437.
Cederman, Lars-Erik, Andreas Wimmer, & Brian Min. 2010. “Why do
Ethnic Groups Rebel? New Data and Analysis.” World Politics.
62 (1), 87–119.
Centlivres, Pierre & Micheline Centlivres-Demont. 1988. “The Afghan
Refugees in Pakistan: A Nation in Exile.” Current Sociology. 36 (2),
71–92.
Chabal, Patrick & Jean-Pascal Daloz. 1999. Africa Works: Disorder as
Political Instrument. Oxford/Bloomington: James Currey/Indiana University Press.
Bibliography
265
Chandra, Kanchan. 2005. “Ethnic Parties and Democratic Stability.” Perspectives on Politics. 3 (2), 235–252.
Checkel, Jeffrey T. 1997. Ideas and International Political Change: Soviet/
Russian Behavior and the End of the Cold War. New Haven: Yale
University Press.
1999. “Norms, Institutions, and National Identity in Contemporary
Europe.” International Studies Quarterly. 43 (1), 83–114.
2001. “Why Comply? Social Learning and European Identity Change.”
International Organization. 55 (3), 553–588.
2003. “‘Going Native’ in Europe? Theorizing Social Interaction in European Institutions.” Comparative Political Studies. 36 (1)–(2), 209–231.
2005. “International Institutions and Socialization in Europe: Introduction and Framework.” International Organization. 59 (4), 801–826.
2006. “Tracing Causal Mechanisms.” International Studies Review. 8 (2),
362–370.
2007. International Institutions and Socialization in Europe. Cambridge
University Press.
2008a. “Process Tracing.” In Audie Klotz & Deepa Prakash (eds.), Qualitative Methods in International Relations: A Pluralist Guide. New
York: Palgrave Macmillan, 114–130.
2008b. “Bridging the Gap? Connecting Qualitative and Quantitative
Methods in the Study of Civil War (Symposium).” Qualitative Methods:
Newsletter of the American Political Science Association Organized
Section for Qualitative and Multi-Method Research. 6 (1), 13–29.
2010. Review of Idean Salehyan, Rebels Without Borders: Transnational
Insurgencies in World Politics. Perspectives on Politics. 8 (4),
1275–1276.
2011. “The Social Dynamics of Civil War: Insights from Constructivist
Theory.” Simons Papers in Security and Development, No.10. Vancouver: School for International Studies, Simon Fraser University (March).
Chernoff, Fred. 2002. “Scientific Realism as a Meta-Theory of International
Relations.” International Studies Quarterly. 46 (2), 189–207.
2007. “Critical Realism, Scientific Realism, and International Relations
Theory.” Millenium: Journal of International Studies. 35 (2), 399–407.
Civil Georgia. 2003. “Security Ministry Unveils Classified Details on Pankisi.”
January 20. www.civil.ge/eng/article.php?id 3033. (Accessed June 22, 2010)
Cline, Lawrence E. 2003. “Spirits and the Cross: Religiously Based Violent
Movements in Uganda.” Small Wars & Insurgencies. 14, 113–130.
Coalition to Stop the Use of Child Soldiers. 2001. “Child Soldiers: Global
Report 2001.” www.child-soldiers.org/global report reader.php?
id 280. (Accessed May 2, 2010)
266
Bibliography
2008. “Child Soldiers: Global Report 2008.” www.childsoldiersglobalreport.
org/files/country pdfs/FINAL 2008 Global Report.pdf. (Accessed October
7, 2008)
Cochrane, Feargal. 2012. “Mediating the Diaspora Space: Charting the
Changing Nature of Irish-America in the Global Age.” In Terence Lyons &
Peter Mandaville (eds.), Politics from Afar: Transnational Diasporas and
Networks. London: Hurst, 117–138.
Cohen, Dara Kay. 2010. “Explaining Sexual Violence During Civil War.”
Ph.D. Dissertation, Stanford University.
Cohen, Robin. 1997. Global Diasporas: An Introduction. London: UCL
Press.
Coleman, James S. 1986. “Social Theory, Social Research, and a Theory of
Action.” The American Journal of Sociology. 91, 1309–1335.
Collier, David. 2011. “Understanding Process Tracing.” PS: Political Science
and Politics. 44, 823–830.
Collier, Paul. 2000. “Economic Causes of Civil Conflict and Their Implications for Policy.” Washington, DC: World Bank Working Paper.
Collier, Paul & Anke Hoeffler. 1998. “On Economic Causes of Civil War.”
Oxford Economic Papers. 50 (4), 563–573.
2004. “Greed and Grievance in Civil War.” Oxford Economic Papers. 56 (4).
Conciliation Resources. 2006. Coming Home. Understanding Why Commanders of the Lord’s Resistance Army Choose to Return to Civilian
Life. London: Conciliation Resources/Quaker Peace & Social Justice.
2010. Initiatives to End the Violence in Northern Uganda: 2002–2009
and the Juba Peace Process. Vol. 11, Accord: An International Review
of Peace Initiatives. London: Conciliation Resources.
Cooley, Alexander & James Ron. 2002. “The NGO Scramble: Organizational Insecurity and the Political Economy of Transnational Action.”
International Security. 27 (1), 5–39.
Cunningham, David, Kristian Gleditsch, & Idean Salehyan. 2006. “Transnational Linkages and Civil War Interactions.” Paper presented at
the Annual Convention of the American Political Science Association,
Philadelphia (September).
2009. “It Takes Two: a Dyadic Analysis of Civil War Duration and
Outcome.” Journal of Conflict Resolution. 53 (4), 570–597.
Cunningham, Kathleen Gallagher. 2011. “Divide and Conquer or Divide
and Concede: How Do States Respond to Internally Divided Separatists?” American Political Science Review. 105 (2), 275–297.
Cunningham, Kathleen, Kristin M. Bakke, & Lee J. M. Seymour. 2012.
“Shirts Today, Skins Tomorrow: Dual Contests and the Effects of
Fragmentation in Self-Determination Conflicts.” Journal of Conflict
Resolution. 56 (1), 67–93.
Bibliography
267
Dalman, Metin & Ismail Tabak. 1995. Avrupa’da Insan Ticareti ve PKK.
Istanbul: Turk Alman Basin Ajansi.
Danforth, Loring M. 1995. The Macedonian Conflict: Ethnic Nationalism in
a Transnational World. Princeton University Press.
Davidson, Jason. 2011. America’s Allies and War: Kosovo, Afghanistan, and
Iraq. New York: Palgrave Macmillan.
de Temmermann, Els. 2001. Aboke Girls. Children Abducted in Northern
Uganda. Kampala: Fountain Press.
della Porta, Donatella. 1996. “Social Movements and the State: Thoughts on
the Policing of Protest.” In Doug McAdam, John D. McCarthy, &
Mayer N. Zald (eds.), Comparative Perspectives on Social Movements:
Political Opportunities, Mobilizing Structures, and Cultural Framings.
New York: Cambridge University Press, 62–92.
della Porta, Donatella & Sidney Tarrow. 2004. “Transnational Processes
and Social Activism: An Introduction.” In Sidney Tarrow & Donnatella
della Porta (eds.), Transnational Protest and Global Activism. Lanham:
Rowman and Littlefield.
Dobbin, Frank, Beth Simmons, & Geoffrey Garrett. 2007. “The Global
Diffusion of Public Policies: Social Construction, Coercion, Competition, or Learning?” Annual Review of Sociology. 33, 449–472.
Dolan, Chris. 2009. Social Torture: The Case of Northern Uganda, 1986–
2006. New York: Berghahn Books.
Doom, Ruddy & Koen Vlassenroot. 1999. “Kony’s Message: A New Koine?
The Lord’s Resistance Army in Northern Uganda.” African Affairs. 98,
5–36.
Dorronsoro, Gilles. 2005. Revolution Unending: Afghanistan, 1979 to the
Present. New York: Columbia University Press.
Dudayev, Umalt. 2004. “Chechnya’s Homemade Weapons Fuel War.” Caucasus Reporting Service, April 15.
Dufoix, Stephane. 2008. Diasporas. Berkeley: University of California Press.
Dunlop, John B. 1998. Russia Confronts Chechnya: Roots of a Separatist
Conflict. New York: Cambridge University Press.
Earl, Jennifer. 2008. “An Admirable Call to Improve, But Not Fundamentally Change, Our Collective Methodological Practices.” Qualitative
Sociology. 31 (4), 355–359.
Egadu, Samuel Richard. 2009. Uganda Probes Alleged Supplies to LRA.
Washington, DC: Institute for War & Peace Reporting.
Ehrenreich, Rosa. 1998. “The Stories We Must Tell: Ugandan Children
and the Atrocities of the Lord’s Resistance Army.” Africa Today. 45,
79–102.
Eccarius-Kelly, Vera. 1999. “Inter-Community Ethnic Conflict in PostUnification Germany: the Kurdish Conundrum.” Paper prepared
268
Bibliography
for the 39th International Studies Association Annual Meeting.
Washington, DC (February).
Elkins, Z., & B. Simmons. 2005. “On Waves, Clusters, and Diffusion:
A Conceptual Framework.” Annals of the American Academy of Political and Social Science. 598 (1), 33–51.
Elman, Colin. 2005. “Explanatory Typologies in Qualitative Studies of
International Politics.” International Organization. 59, 293–326.
Elman, Colin & Miriam Elman. 2003. “Introduction,” and “Lessons From
Lakatos.” In Colin Elman & Miriam Elman (eds.), Progress in International
Relations Theory: Appraising the Field. Cambridge, MA: MIT Press, 1–70.
Elster, Jon. 1998. “A Plea for Mechanisms.” In Peter Hedstroem & Richard
Swedberg (eds.), Social Mechanisms: An Analytical Approach to Social
Theory. Cambridge University Press, 45–73.
Entessar, Nader. 1992. Kurdish Ethnonationalism. Boulder, CO: Lynne
Rienner.
Epstein, J. M., & R. Axtell. 1996. Growing Artificial Societies: Social Science
from the Bottom Up. Washington, DC: Brookings Institution Press.
Evangelista, Matthew. 1995. “The Paradox of State Strength: Transnational
Relations, Domestic Structures, and Security Policy in Russia and the
Soviet Union.” International Organization. 49 (1), 1–38.
1999. Unarmed Forces: The Transnational Movement to End the Cold
War. Ithaca, NY: Cornell University Press.
2002. The Chechen Wars: Will Russia Go the Way of the Soviet Union?
Washington, DC: Brookings Institution Press.
Fair, Christine C. 2005. “Diaspora Involvement in Insurgencies: Insights
from the Khalistan and Tamil Eelam Movements.” Nationalism and
Ethnic Politics. 11 (1), 125–156.
2009. The Madrasa Challenge: Militancy and Religious Education in
Pakistan. Washington, DC: United States Institute of Peace.
Falleti, Tulia & Julia Lynch. 2009. “Context and Causal Mechanisms in
Political Analysis.” Comparative Political Studies. 42 (9), 1143–1166.
Farmar, Sam. 2006. “Uganda Rebel Leader Breaks Silence.” BBC Newsnight, June 28.
Fearon, James & David Laitin. 2003. “Ethnicity, Insurgency and Civil War.”
American Political Science Review. 97 (1), 75–90.
Fielden, Matthew B. 1998. “The Geopolitics of Aid: The Provision and
Termination of Aid to Afghan Refugees in North-West Frontier Province, Pakistan.” Political Geography. 17 (4), 459–487.
Finnemore, Martha & Kathryn Sikkink. 1998. “International Norm Dynamics and Political Change.” International Organization. 52, 887–917.
Bibliography
269
Finnström, Sverker. 2008. Living With Bad Surroundings: War, History,
and Everyday Moments in Northern Uganda. Durham: Duke University
Press.
Forsberg, E. 2008. “Polarization and Ethnic Conflict in a Widened Strategic
Setting.” Journal of Peace Research. 45 (2), 283–300.
Fortna, Virginia Page. 2004. “Interstate Peacekeeping: Causal Mechanisms
and Empirical Effects.” World Politics. 56 (4), 481–519.
Franco, Claudio. 2009. “The Tehrik-e Taliban Pakistan.” In Antonio
Giustozzi (ed.), Decoding the New Taliban: Insights from the Afghan
Field. New York: Columbia University Press, 269–291.
Franzese, Robert & Jud Hays. 2008. “Interdependence in Comparative
Politics: Substance, Theory, Empirics, Substance.” Comparative Political Studies. 41 (4/5), 742–780.
Friedman, Milton. 1953. “The Methodology of Positive Economics.”
Reprinted in Uskali Maki (ed.), The Methodology of Positive Economics: Reflections on the Milton Friedman Legacy. Cambridge University
Press (2009). 3–44.
Frist, William, H. 2009. A Heart to Serve: The Passion to Bring Health,
Hope and Healing. Nashville, TN: Center Street Publishing.
Fujii, Lee Ann. 2009. Killing Neighbors: Webs of Violence in Rwanda.
Ithaca, NY: Cornell University Press.
2010. “Shades of Truth and Lies: Interpreting Testimonies of War and
Violence.” Journal of Peace Research. 47 (2), 231–241.
Gagnon, V. P. Jr. 1994. “Ethnic Nationalism and International Conflict: The
Case of Serbia.” International Security. 19 (3), 130–166.
Gall, Carlotta & Thomas de Waal. 1998. Chechnya: Calamity in the Caucasus. New York University Press.
Gammer, Moshe. 2005. “Between Mecca and Moscow: Islam, Politics,
and Political Islam in Chechnya and Daghestan.” Middle Eastern
Studies. 41 (6), 833–848.
2006. The Lone Wolf and the Bear: Three Centuries of Chechen Defiance
of Russian Rule. Pittsburg University Press.
Gamson, William A. 1975. The Strategy of Social Protest. Homewood: The
Dorsey Press.
Gates, Scott. 2002. “Recruitment and Allegiance: The Microfoundations of
Rebellion.” Journal of Conflict Resolution. 46 (1), 111–130.
2008. “Mixing it Up: The Role of Theory in Mixed-Methods Research.”
Qualitative Methods: Newsletter of the American Political Science
Association Organized Section for Qualitative and Multi-Method
Research. 6 (1), 27–29.
270
Bibliography
2011. “The Dynamics of Group Allegiance.” Memo prepared for a workshop on “The Social Dynamics of Civil War.” Peace Research Institute
Oslo, Centre for the Study of Civil War (February).
Gates, Scott & Simon Reich (eds). 2009. Child Soldiers in the Age of
Fractured States. Pittsburg University Press.
George, Alexander. 1993. Bridging the Gap: Theory and Practice in Foreign
Policy. Washington, DC: United States Institute of Peace Press.
George, Alexander & Andrew Bennett. 2005. Case Studies and Theory
Development in the Social Sciences. Cambridge, MA: The MIT Press.
German, Tracey C. 2003. Russia’s Chechen War. New York: Routledge
Curzon.
Gerring, John. 2007a. Case Study Research: Principles and Practices. Cambridge University Press.
2007b. “The Mechanismic Worldview: Thinking Inside the Box.” British
Journal of Political Science. 38, 161–179.
Gerring, John & Paul Barresi. 2003. “Putting Ordinary Language to Work:
A Min-Max Strategy of Concept Formation in the Social Sciences.”
Journal of Theoretical Politics. 15 (2), 20–32.
Gersony, Robert 1997. The Anguish of Northern Uganda. Results of a Fieldbased Assessment of the Civil Conflicts in Northern Uganda. Kampala:
USAID Mission.
Gheciu, Alexandra. 2005. “Security Institutions as Agents of Socialization?
NATO and Post-Cold War Central and Eastern Europe.” International
Organization. 59 (4), 973–1012.
Giddens, Anthony. 1984. The Constitution of Society: Outline of a Theory
of Structuration. Cambridge: Polity Press.
Gilardi, Fabrizio. 2010. “Who Learns from What in Policy Diffusion Processes.” American Journal of Political Science. 54 (3), 650–666.
Giugni, Marco. 1999. “How Social Movements Matter: Past Research, Present Problems, Future Developments.” In Marco Giugni, Doug McAdam, & Charles Tilly (eds.), How Social Movements Matter. University
of Minnesota Press, xiii–xxxiii.
Giuliano, Elise. 2000. “Who Determines the Self in the Politics of SelfDetermination? Identity and Preference Formation in Tatarstan’s
Nationalist Mobilization.” Comparative Politics. 32, 295–316.
Giustozzi, Antonio. 2007. Koran, Kalashnikov, Laptop: The Neo-Taliban
Insurgency in Afghanistan. London: Hurst.
2009. Empires of Mud: Wars and Warlords in Afghanistan. London:
Hurst & Co.
Bibliography
271
Gleditsch, Kristian Skrede. 2002. All International Politics is Local: The
Diffusion of Conflict, Integration, and Democratization. Ann Arbor:
University of Michigan Press.
2007. “Transnational Dimensions of Civil War.” Journal of Peace
Research. 44 (3), 293–309.
Gleditsch, Kristian Skrede and Idean Salehyan. 2006. “Refugees and
the Spread of Civil War.” International Organization. 60 (2),
335–366.
Goddard, Stacie. 2007. “When Position is Power: Entrepreneurs, Networks
and International Politics.” Paper prepared for delivery at the Annual
Convention of the American Political Science Association, Chicago.
Goertz, Gary & James Mahoney. 2004. “The Possibility Principle: Choosing
Negative Cases in Comparative Research.” American Political Science
Review. 98 (4), 653–659.
Goffman, Erving. 1974. Frame Analysis: An Essay on the Organization of
Experience. London: Harper and Row.
Goldstone, Jack. 2002. “Population and Security: How Demographic
Change can Lead to Violent Conflict.” Columbia Journal of International Affairs 56, 245–263.
Goodwin, Jeff. 2001. No Other Way Out: States and Revolutionary Movements 1945–1991. New York: Cambridge University Press.
Gopal, Anand. 2010. “The Battle for Afghanistan.” Washington, DC: New
America Foundation.
Griswold, Eliza. 2008. “The Man for a New Sudan.” New York Times Magazine. June 15. www.nytimes.com/2008/06/15/magazine/15SUDAN-t.html?
pagewanted all. (Accessed May 15, 2010)
Gunaratna, Rohan. 2001. “Dynamics of Contemporary Terrorist and
Guerrilla Campaigns: International and Domestic Factors Facilitating
and Inhibiting PKK, LTTE, ETA, FARC, GIA and MTA to Escalate
and De-escalate Violence.” Paper Prepared for the Workshop, “Trajectories of Terrorist Violence in Europe” at the Minda de Gunzburg Center
for European Studies, Harvard University.
Gunter, Michael. 1990. The Kurds in Turkey: A Political Dilemma. Boulder,
CO: Westview Press.
Gusterson, Hugh. 2008. “Ethnographic Research.” In Audie Klotz & Deepa
Prakash (eds.), Qualitative Methods in International Relations:
A Pluralist Guide. New York: Palgrave Macmillan, 93–113.
Haas, Peter. 1992. “Introduction: Epistemic Communities and International
Policy Coordination.” International Organization. 46 (1), 1–35.
272
Bibliography
Hafner-Burton, Emily. 2008. “Sticks and Stones: Naming and Shaming the
Human Rights Enforcement Problem.” International Organization.
62 (4), 689–716.
2009. Forced to be Good: Why Trade Agreements Boost Human Rights.
Ithaca, NY: Cornell University Press.
Hafner-Burton, Emily & James Ron. 2009. “Seeing Double: Human Rights
Impact through Qualitative and Quantitative Eyes.” World Politics.
61 (2), 360–401.
Haider, Ziad. 2005. “Sino-Pakistan Relations and Xinjiang’s Uighurs: Politics, Trade, and Islam along the Karakoram Highway.” Asian Survey. 45,
522–545.
Hall, Peter. 2003. “Aligning Ontology & Methodology in Comparative
Politics.” In James Mahoney & Dietrich Rueschemeyer (eds.), Comparative Historical Analysis in the Social Sciences. Cambridge University Press, 373–406.
Harpviken, Kristian Berg. 1996. “Political Mobilization among the Hazara of
Afghanistan: 1978–1992.” University of Oslo, Department of Sociology.
1997. “Transcending Traditionalism: The Emergence of Non-State Military Formations in Afghanistan.” Journal of Peace Research. 34 (3),
271–287.
2009. Social Networks and Migration in Wartime Afghanistan. Houndmills:
Palgrave Macmillan.
2010. “The Return of the Refugee Warrior: Migration and Armed Resistance in Herat.” In Ceri Oeppen & Angela Schlenkhoff (eds.), Beyond
the ‘Wild Tribes’: Understanding Modern Afghanistan and Its Diaspora. London: Hurst, 121–137.
Hassanpour, Amir. 1998. “Satellite Footprints as National Borders: MEDTV and the Extraterritoriality of State Sovereignty.” Journal of Muslim
Minority Affairs. 18 (1), 53–72.
Hedström, P. & R. Swedberg (eds.). 1998. Social Mechanisms: An Analytical
Approach to Social Theory. New York: Cambridge University Press.
Hegghammer, Thomas. 2010. “The Rise of Muslim Foreign Fighters: Islam
and the Globalization of Jihad.” International Security. 35 (3), 53–94.
Hegre, Håvard & Nicholas Sambanis. 2006. “Sensitivity Analysis of Empirical Results on Civil War Onset.” Journal of Conflict Resolution. 50 (4),
508–535.
Helfer, Hans-Ulrich. 1988. Arbeiterpartei Kurdistans (PKK) – Organisation –
Aktivitaeten in der Schweiz. Zurich: Presdok.
Hempel, Carl & P. Oppenheim. 1948. “Studies in the Logic of Explanation.”
Philosophy of Science 15: 135–175. Reprinted in Carl Hempel. [1965].
Aspects of Scientific Explanation and Other Essays in the Philosophy of
Science. New York: Free Press, 245–290.
Bibliography
273
Henkin, Yagil. 2006. “From Tactical Terrorism to Holy War: The Evolution
of Chechen Terrorism.” Central Asian Survey. 25 (1), 193–203.
Hertzke, Allen D. 2004. Freeing God’s Children: The Unlikely Alliance for
Global Human Rights. New York: Rowman & Littlefield.
Hill, S. & D. Rothchild. 1986. “The Contagion of Political Conflict in Africa
and the World.” Journal of Conflict Resolution. 30 (4), 716–735.
Hironaka, Ann. 2005. Neverending Wars: The International Community,
Weak States, and the Perpetuation of Civil War. Cambridge, MA:
Harvard University Press.
Ho, Enseng. 2004. “Empire Through Diasporic Eyes: A View from the Other
Boat.” Comparative Studies in Society and History. 46 (2), 17–44.
Hockenos, Paul. 2003. Homeland Calling: Exile Patriotism and the Balkan
Wars. Ithaca, NY: Cornell University Press.
Hoffmann, Matthew. 2008. “Agent-based Modeling.” In Audie Klotz &
Deepa Prakash (eds.), Qualitative Methods in International Relations:
A Pluralist Guide. New York: Palgrave Macmillan, 187–210.
Hopf, Ted. 2007. “The Limits of Interpreting Evidence.” In Mark
Lichbach & Ned Lebow (eds.), Theory and Evidence in Comparative
Politics and International Relations. New York: Palgrave Macmillan,
55–86.
Horowitz, Donald. 1985. Ethnic Groups in Conflict. University of California
Press.
Hughes, James. 2007. Chechnya: From Nationalism to Jihad. University of
Pennsylvania Press.
Human Rights Watch. 1994. “Sudan: The Lost Boys.” New York: Human
Rights Watch.
1996. “Children in Combat.” New York: Human Rights Watch.
1997. “The Scars of Death. Children Abducted by the Lord’s Resistance
Army in Uganda.” New York: Human Rights Watch.
2005. “Uprooted and Forgotten: Impunity and Human Rights Abuses in
Northern Uganda.” New York: Human Rights Watch.
2006. “Sudan: Regional Government Pays Ugandan Rebels not to
Attack.” New York: Human Rights Watch.
Humphreys, Macartan, & Jeremy M. Weinstein. 2006. “Handling and
Manhandling Civilians in Civil War.” American Political Science
Review. 100 (3), 429–447.
2007. “Demobilization and Reintegration.” Journal of Conflict Resolution. 51 (4), 531–567.
Huntington, Samuel P. 1996. The Clash of Civilizations and the Remaking
of World Order. New York: Simon and Schuster.
1997. “The Erosion of American National Interest.” Foreign Affairs.
76 (5), 28–49.
274
Bibliography
Imset, Ismet G. 1992. The PKK: A Report on Separatist Violence in Turkey.
Ankara: Turkish Daily News Publications.
International Crisis Group. 2002. “God, Oil and Country: Changing the
Logic of War in Sudan.” ICG Africa Report no. 39. www.crisisgroup.org/
library/documents/report archive/A400534 28012002.pdf. (Accessed
February 19, 2008)
2003. “Afghanistan: The Problem of Pashtun Alienation.” ICG Asia
Report. Brussels/Kabul: International Crisis Group.
2004. “Northern Uganda: Understanding and Solving the Conflict.” ICG
Africa Report no.77. Brussels/Nairobi: International Crisis Group.
2010. “LRA: A Regional Strategy Beyond Killing Kony.” ICG Africa
Report no. 157. Brussels/Nairobi: International Crisis Group.
Izady, Mehrdad. 1992. The Kurds: A Concise History. London: Taylor and
Francis.
Jackson, Paul. 2002. “The March of the Lord’s Resistance Army. Greed
or Grievance in Northern Uganda?” Small Wars & Insurgencies. 13,
29–52.
James, Paul & R. R. Sharma. 2006. “Transnational Conflict: A Critical
Introduction.” In Paul James & R. R. Sharma (eds.), Globalization
and Violence: vol. IV, Transnational Conflict. London: Sage, ix–xxxiv.
Jamestown Foundation. 1995. “Prism Interview: ‘We Set Off For Budennovsk with a One-Way Ticket’.” Prism. October 20.
1996. “Peer Criticizes Performance of Chechen Commander Raduyev.”
Monitor. January 26.
2006. “Basayev Issues Statement; Umarov Gives Interview.” Chechnya
Weekly. April 20.
2007. “Chechens Weigh In on the ‘Caucasus Emirate’ Idea.” North Caucasus Analysis. November 8.
Jenne, Erin. 2007. Ethnic Bargaining: The Paradox of Minority Empowerment. Ithaca, NY: Cornell University Press.
Johnson, Douglas, H. 1989. “The Structuring of a Legacy: Military Slavery
in Northeast Africa.” Ethnohistory. 36 (1), 72–88.
1992. “Recruitment and Entrapment in Private Slave Armies: The Structure
of Zara’ib in the Southern Sudan.” Slavery & Abolition. 13 (1), 162–173.
2003. The Root Causes of Sudan’s Civil Wars. Indiana University Press.
Johnson, James. 2006. “Consequences of Positivism: A Pragmatist Assessment.” Comparative Political Studies. 39 (2), 224–252.
Johnson, Scott. 2009. “Hard Target. The Hunt for Africa’s Last Warlord.”
Newsweek. May 16.
Johnston, Alastair Iain. 2001. “Treating International Institutions as Social
Environments.” International Studies Quarterly 45/4: 487–516.
Bibliography
275
2005. “Conclusions and Extensions: Toward Mid-Range Theorizing and
Beyond Europe.” International Organization. 59 (4), 1013–1044.
2008. Social States: China in International Institutions, 1980–2000.
Princeton University Press.
Justice and Peace Commission of the Archdiocese of Gulu. 2008. “Northern
Situation Now.” Justice and Peace News. 8, 2–4.
Kaldor, Mary. 1999. New and Old Wars: Organized Violence in a Global
Era. Stanford University Press.
Kalyvas, Stathis. 2003. “The Ontology of ‘Political Violence’: Action and
Identity in Civil Wars.” Perspectives on Politics. 1, 475–494.
2006. The Logic of Violence in Civil War. Cambridge University Press.
2007. Review of Jeremy Weinstein, Inside Rebellion: The Politics of Insurgent Violence, in Comparative Political Studies. 40 (9), 1146–1151.
2008. “Ethnic Defection in Civil War.” Comparative Political Studies.
41 (8), 1043–1068.
Kalyvas, Stathis & Laia Balcells. 2010. “International System and Technologies of Rebellion: How the End of the Cold War Shaped Internal
Conflict.” American Political Science Review. 104 (3), 415–429.
Kastner, Philipp. 2007. “ICC in Darfur: Savior or Spoiler?” ILSA Journal of
International & Comparative Law. 14, 145–188.
Katzenstein, Peter, and Nobuo Okawara. 2002. “Japan, Asian-Pacific Security, and the Case for Analytical Eclecticism.” International Security.
26 (3), 153–185.
Katzenstein, Peter & Rudra Sil. 2008. “Eclectic Theorizing in the Study and
Practice of International Relations.” In Christian Reus-Smit and
Duncan Snidal (eds.), The Oxford Handbook of International Relations. New York: Oxford University Press, 109–130.
Katzenstein, Suzanne & Jack Snyder. 2009. “Expediency of Angels.” The
National Interest. March–April, 58–65.
Keck, Margaret & Kathryn Sikkink. 1998. Activists Beyond Borders: Advocacy Networks in International Politics. Ithaca, NY: Cornell University
Press.
Keohane, Robert and Joseph Nye (eds.). 1972. Transnational Relations and
World Politics. Harvard University Press.
Khagram, Sanjeev, James V. Riker & Kathryn Sikkink (eds.). 2002. Restructuring World Politics: Transnational Social Movements, Networks, and
Norms. Minneapolis: University of Minnesota Press.
King, Charles. 2004. “The Micropolitics of Social Violence (A Review
Essay).” World Politics. 56 (3), 431–455.
2008. The Ghost of Freedom: A History of the Caucasus. New York:
Oxford University Press.
276
Bibliography
King, Charles & Neil J. Melvin. 2000. “Diaspora Politics: Ethnic Linkages,
Foreign Policy, and Security in Eurasia.” International Security. 24 (3),
108–138.
King, Gary, Robert Keohane, & Sidney Verba. 1994. Designing Social
Inquiry. Princeton University Press.
Kinzer, Stephen. 2008. A Thousand Hills: Rwanda’s Rebirth and the Man
Who Dreamed It. New Jersey: John Wiley.
Klotz, Audie. 1995. Norms in International Relations: The Struggle Against
Apartheid. Ithaca, NY: Cornell University Press.
Koinova, Maria. 2009. “Diasporas and Democratization in the PostCommunist World.” Communist and Post-Communist Studies. 42,
41–64.
2010. “Diasporas and Secessionist Conflicts: The Mobilization of the
Armenian, Albanian and Chechen Diasporas. Ethnic and Racial Studies. 34 (2), 333–356.
Kuehn, David & Ingo Rohlfing. 2009. “Does It, Really? Measurement Error
and Omitted Variables in Multi-Method Research.” Qualitative
Methods: Newsletter of the American Political Science Association
Organized Section for Qualitative and Multi-Method Research. 7 (2),
18–22.
Kuhn, Thomas. 1962. The Structure of Scientific Revolutions. University of
Chicago Press.
Kuperman, Alan J. 2004. “Provoking Genocide: A Revised History of the
Rwandan Patriotic Front.” Journal of Genocide Research. 6 (1), 61–84.
Kuran, T. 1998. “Ethnic Dissimilation and Its International Diffusion.” In
D. A. Lake & D. Rothchild (eds.), The International Spread of Ethnic
Conflict: Fear, Diffusion, and Escalation. Princeton University Press,
35–60.
Kurki, Milja. 2008. Causation in International Relations: Reclaiming Causal
Analysis. Cambridge University Press.
Kydd, Andrew & Barbara Walter. 2006. “Strategies of Terrorism.” International Security. 31 (1), 49–80.
Lakatos, Imre. 1970. “Falsification and the Methodology of Scientific
Research Programmes.” In Imre Lakatos & A. Musgrave (eds.), Criticism and the Growth of Knowledge. Cambridge, NY: Cambridge University Press, 91–196.
Landesamt für Verfassungsschutz Berlin. 1994. Auslaenderextremismus in
Berlin. Berlin: Landesamt für Verfassungsschutz.
Lawrence, Adria. 2010. “Triggering Nationalist Violence: Competition and
Conflict in Uprisings against Colonial Rule.” International Security.
35 (2), 88–122.
Bibliography
277
Leggewie, C. 1996. “How Turks Became Kurds, Not Germans.” Dissent. 43,
79–83.
Leiby, Michele. 2009a. “Digging in the Archives: The Promise and Perils of
Primary Documents.” Politics & Society. 37 (1), 75–100.
2009b. “Wartime Sexual Violence in Guatemala and Peru.” International
Studies Quarterly. 53 (2), 445–468.
Lemarchand, René. 2001. “Exclusion, Marginalization and Political Mobilization: The Road to Hell in the Great Lakes.” CAS Africa Papers.
Centre of African Studies, University of Copenhagen.
Levitt, Peggy & Sally Merry. 2009. “Vernacularization on the Ground:
Local uses of Global Women’s Rights in Peru, China, India and the
United States.” Global Networks. 9 (4), 441–461.
Levy, Marc, Oran Young, & Michael Zuern. 1995. “The Study of International Regimes.” European Journal of International Relations. 1 (3),
267–330.
Lewis, Jeffrey. 2005. “The Janus Face of Brussels: Socialization and Everyday Decision Making in the European Union.” International Organization. 59 (4), 937–972.
Lichbach, Mark I. 1994. “What Makes Rational Peasants Revolutionary?
Dilemma, Paradox, and Irony in Peasant Collective Action.” World
Politics. 46 (3), 383–418.
1998. The Rebels’ Dilemma: Economics, Cognition, and Society. Ann
Arbor, University of Michigan Press.
2008. “Modeling Mechanisms of Contention: MTT’s Positivist Constructivism.” Qualitative Sociology. 31 (4), 345–354.
Lischer, Sarah Kenyon. 2003. “Collateral Damage: Humanitarian Assistance
as a Cause of Conflict.” International Security. 28 (1), 79–109.
2005. Dangerous Sanctuaries: Refugee Camps, Civil Wars and the
Dilemmas of Humanitarian Aid. Ithaca, NY: Cornell University Press.
Loescher, Gil. 1992. Refugee Movements and International Security.
London: International Institute for Strategic Studies.
Lynch, Marc. 2002. “Why Engage? China and the Logic of Communicative
Engagement.” European Journal of International Relations. 8 (2), 187–230.
Lyon, Alynna J. & Emek M. Ucarer. 2001. “Mobilizing Ethnic Conflict:
Kurdish Separatism in Germany and the PKK.” Ethnic and Racial
Studies. 24 (6), 925–948.
Lyons, Terrence. 2006. “Diasporas and Homeland Conflict.” In Miles
Kahler & Barbara Walter (eds.), Territoriality and Conflict in an Era
of Globalization. New York: Cambridge University Press, 111–131.
Mahoney, James. 2000. “Path Dependence in Historical Sociology.” Theory
and Society. 29 (4), 507–538.
278
Bibliography
2001. “Review – Beyond Correlational Analysis: Recent Innovations in
Theory and Method.” Sociological Forum. 16 (3), 575–593.
Malet, David. 2007. “‘The More Irregular the Service’: Transnational Identity
Communities and the Foreign Fighter.” Paper presented at the annual
meeting of the American Political Science Association, Chicago, September.
2010. “Why Foreign Fighters? Historical Perspectives and Solutions.”
Orbis. 54 (1), 97–114.
2011. “The Foreign Fighter Project.” http://davidmalet.com/The Foreign
Fighter Project.php. (Accessed September 7, 2010)
Maliniak, Daniel, Amy Oakes, Susan Peterson, & Michael Tierney. 2011.
“International Relations in the US Academy,” International Studies
Quarterly, 437–464.
Malkki, Lisa. 1995. “Refugees and Exile: From ‘Refugee Studies’ to the
National Order of Things.” Annual Review of Anthropology. 24,
495–523.
Mamdani, Mahmood. 1984. Imperialism and Fascism in Uganda. Trenton,
NJ: Africa World Press.
2001. When Victims Become Killers: Colonialism, Nativism, and the
Genocide in Rwanda. Princeton University Press.
Mantilla, Giovanni. 2009. “Emerging International Human Rights Norms
for Transnational Corporations.” Global Governance. 15, 279–298.
March, James & Johan Olsen. 1984. “The New Institutionalism: Organizational Factors in Political Life.” American Political Science Review.
78 (3), 734–749.
Marcus, Aliza. 1993. “Turkey’s Kurds After the Gulf War: A Report from
the Southeast.” In Gerard Chaliand (ed.), A People Without a Country:
The Kurds and Kurdistan. London: Zed Press.
Markides, Kyriacos C. 1977. The Rise and Fall of the Cyprus Republic. New
Haven, Conn.: Yale University Press.
Mathias, Charles, Jr. 1981. “Ethnic Groups and Foreign Policy.” Foreign
Affairs. 59, 975–998.
Mayntz, Renate. 2004. “Mechanisms in the Analysis of Macro-Social
Phenomena.” Philosophy of the Social Sciences. 34 (2), 237–259.
McAdam, Doug. 1983. “Tactical Innovation and the Pace of Insurgency.”
American Sociological Review. 48 (6), 735–754.
McAdam, Doug, John D. McCarthy, & Mayer N. Zald (eds.). 1996. Comparative Perspectives on Social Movements: Political Opportunities,
Mobilizing Structures, and Cultural Framings. New York: Cambridge
University Press.
McAdam, Doug & Sidney Tarrow. 2005. “Scale Shift and Transnational
Contention.” In Sidney Tarrow & Donnatella della Porta (eds.),
Bibliography
279
Transnational Protest and Global Activism. Lanham, MD: Rowman
and Littlefield, 121–150.
McAdam, Doug, Sidney Tarrow, & Charles Tilly. 1996. “To Map Contentious Politics.” Mobilization: An International Journal. 1 (1), 17–34.
2001. Dynamics of Contention. Cambridge University Press.
2008. “Methods for Measuring Mechanisms of Contention.” Qualitative
Sociology. 31 (4), 307–331.
McCarthy, John D. & Mayer N. Zald. 1977. “Resource Mobilization and
Social Movements: A Partial Theory.” American Journal of Sociology.
82 (6), 1212–1241.
McDowall, David. 1996. A Modern History of the Kurds. London: IB
Tauris.
McGrory, Mary. 2001. “Suddenly, Sudan.” Washington Post. March 11.
McKeown, Timothy. 1999. “Case Studies and the Statistical Worldview:
Review of King, Keohane and Verba’s Designing Social Inquiry: Scientific Inference in Qualitative Research.” International Organization.
53 (1), 161–190.
Meseguer, Covadonga. 2006. “Rational Learning and Bounded Learning in
the Diffusion of Policy Innovation.” Rationality and Society. 18 (1),
35–66.
Meseguer, Covadonga & Fabrizio Gilardi. 2009. “What is New in the Study
of Policy Diffusion?” Review of International Political Economy.
16 (3), 527–543.
Meyer, David S. & Nancy Whittier. 1994. “Social Movement Spillover.”
Social Problems. 41 (2), 277–298.
Mgbako, Chi. 2005. “Ingando Solidarity Camps: Reconciliation and Political Indoctrination in Post-Genocide Rwanda.” Harvard Human Rights
Journal. 18, 201–224.
Miliband, David. 2009. “The Army Alone cannot Defeat this Taliban Insurgency.” The Independent.
Minear, Larry, & Randolph C. Kent. 1998. “Rwanda’s Internally Displaced:
A Conundrum Within a Conundrum.” In Roberta Cohen & Francis M.
Deng (eds.), The Forsaken People: Case Studies of the Internally
Displaced. Washington, DC: The Brookings Institution, 57–95.
Moghaddam, Sippi Azerbaijani, Mirwais Wardak, Idrees Zaman, & Annabel Taylor. 2008. “Afghan Hearts, Afghan Minds: Exploring Afghan
Perceptions of Civil-Military Relations.” Kabul: European Network of
NGOs in Afghanistan (ENNA) and the British and Irish Agencies
Afghanistan Group (BAAG).
Moore, Cerwyn. 2007. “The Radicalisation of the Chechen Separatist
Movement – Myth or Reality?” Prague Watchdog. May 16.
280
Bibliography
Moore, Cerwyn, & Paul Tumelty. 2008. “Foreign Fighters and the Case of
Chechnya: A Critical Assessment.” Studies in Conflict and Terrorism.
31, 412–43.
2009. “Assessing the Unholy Alliances in Chechnya: From Communism
and Nationalism to Islamism and Salafism.” Journal of Communist
Studies and Transition Politics. 25 (1), 73–94.
Moravcsik, Andrew. 1993. “Introduction – Integrating International and
Domestic Theories of International Bargaining.” In Peter Evans, Harold
Jacobson, & Robert Putnam (eds.), Double-Edged Diplomacy: International Bargaining and Domestic Politics. Berkeley: University of California Press.
Murphy, Paul J. 2004. The Wolves of Islam: Russia and the Faces of
Chechen Terror. Dulles, VA: Brassey’s.
Museveni, Yoweri Kaguta. 1992. What is Africa’s Problem? Kampala: NRM
Publications.
Muzaev, Timur. 1997. “Chechenskaja Respublika Ichkerija” [The Chechen
Republic of Ichkeria], Political Monitoring, Institute for Humanitarian
and Political Studies (IGPI), October. http://igpi.ru/monitoring/north
caucas. (Accessed May 3, 2011)
1998a. “Chechenskaja Respublika Ichkerija” [The Chechen Republic of
Ichkeria], Political Monitoring, Institute for Humanitarian and Political
Studies (IGPI), January. http://igpi.ru/monitoring/north caucas.
(Accessed May 3, 2011)
1998b. “Chechenskaja Respublika Ichkerija” [The Chechen Republic of
Ichkeria], Political Monitoring, Institute for Humanitarian and Political
Studies (IGPI), April. http://igpi.ru/monitoring/north caucas. (Accessed
May 3, 2011)
1998c. “Chechenskaja Respublika Ichkerija” [The Chechen Republic of
Ichkeria], Political Monitoring, Institute for Humanitarian and Political
Studies (IGPI), June. http://igpi.ru/monitoring/north caucas. (Accessed
May 3, 2011)
1998d. “Chechenskaja Respublika Ichkerija” [The Chechen Republic of
Ichkeria], Political Monitoring, Institute for Humanitarian and Political
Studies (IGPI), November. http://igpi.ru/monitoring/north caucas.
(Accessed May 3, 2011)
1998e. “Chechenskaja Respublika Ichkerija” [The Chechen Republic of
Ichkeria], Political Monitoring, Institute for Humanitarian and Political
Studies (IGPI), December. http://igpi.ru/monitoring/north caucas.
(Accessed May 3, 2011)
1999a. “Chechenskaja Respublika Ichkerija” [The Chechen Republic of
Ichkeria], Political Monitoring, Institute for Humanitarian and Political
Bibliography
281
Studies (IGPI), February. http://igpi.ru/monitoring/north caucas.
(Accessed May 3, 2011.
1999b. “Chechenskaja Respublika Ichkerija” [The Chechen Republic of
Ichkeria], Political Monitoring, Institute for Humanitarian and Political
Studies (IGPI), April. http://igpi.ru/monitoring/north caucas. (Accessed
May 3, 2011)
1999c. “Chechenskaja Respublika Ichkerija” [The Chechen Republic of
Ichkeria], Political Monitoring, Institute for Humanitarian and Political
Studies (IGPI), May. http://igpi.ru/monitoring/north caucas. (Accessed
May 3, 2011)
1999d. “Chechenskaja Respublika Ichkerija” [The Chechen Republic of
Ichkeria], Political Monitoring, Institute for Humanitarian and Political
Studies (IGPI), June. http://igpi.ru/monitoring/north caucas. (Accessed
May 3, 2011)
1999e. “Chechenskaja Respublika Ichkerija” [The Chechen Republic of
Ichkeria], Political Monitoring, Institute for Humanitarian and Political
Studies (IGPI), July. http://igpi.ru/monitoring/north caucas. (Accessed
May 3, 2011)
Neu, Joyce. 2002. “Restoring Relations Between Uganda and Sudan: The
Carter Center Process.” Accord. Protracted Conflict, Elusive Peace. 11,
46–51.
Neumann, Iver. 2008. “Discourse Analysis.” In Audie Klotz & Deepa
Prakash (eds.), Qualitative Methods in International Relations:
A Pluralist Guide. New York: Palgrave Macmillan, 61–77.
Nivat, Anne. 2001. Chienne de Guerre: A Woman Reporter Behind the
Lines of War in Chechnya. New York: Public Affairs.
2005. “The Black Widows: Chechen Women Join the Fight for Independence – and Allah.” Studies in Conflict and Terrorism. 28 (5), 413–419.
Nixon, Hamish. 2011. “Prospects for Durable Peace in Afghanistan:
A Review of the Literature (draft).” Bergen/Oslo/Washington, DC:
Chr. Michelsen Institute/Peace Research Institute Oslo/United States
Institute of Peace.
Nye, Joesph S., & Robert O. Keohane. 1971. “Transnational Relations and
World Politics: An Introduction.” International Organization. 25 (3),
329–349.
Obama, Barack. 2011. Letter from the President to the Speaker of the House
of Representatives and the President Pro Tempore of the Senate
regarding the Lord’s Resistance Army (October 14). Washington, DC:
The White House.
Obita, James Alfred. 1997. “A Case for National Reconciliation, Peace,
Democracy, and Economic Prosperity for all Ugandans.” The official
282
Bibliography
presentation of the Lord’s Resistance Movement/Army (LRA/M).
Kacoke Madit Conference. University of London.
O’Kadameri, Billie. 2002. “LRA/Government Negotiations 1993–4.”
Accord: Protracted Conflict, Elusive Peace. 11, 34–41.
Okiror, Samuel. 2007. “Museveni Wants LRA Warrants - For Now.” ACR
Issue 123. Washington, DC: Institute for War & Peace Reporting.
Olson, Mancur. 1965. The Logic of Collective Action. Cambridge, MA:
Harvard University Press.
Olson, Mancur & Richard Zeckhauser. 1966. “An Economic Theory of
Alliances.” The Review of Economics and Statistics. 48 (3), 266–279.
Omara-Otunnu, Amii. 1987. Politics and the Military in Uganda,
1890–1985. New York: St. Martin’s Press.
1992. “The Struggle for Democracy in Uganda.” The Journal of Modern
African Studies. 30, 443–463.
O’Neill, Bard E. 2001. Insurgency & Terrorism: Inside Modern Revolutionary Warfare. New York: Brassey’s.
Orenstein, Mitchell & Hans Peter Schmitz. 2006. “The New Transnationalism and Comparative Politics.” Comparative Politics. 38 (4), 479–500.
stergaard-Nielsen, Eva. 2001. “Transnational Political Practices and the
Receiving State: Turks and Kurds in Germany and the Netherlands.”
Global Networks: A Journal of Transnational Affairs. 1 (3), 261–282.
2003. Transnational Politics: Turks and Kurds in Germany. London:
Routledge.
Otim, Michael & Marieke Wierda. 2010. Uganda. Impact of the Rome
Statute and the International Criminal Court. New York: International
Center for Transitional Justice.
Padgett, John F. & Christopher K. Ansell. 1993. “Robust Action and the
Rise of the Medici, 1400–1434.” American Journal of Sociology. 98 (6),
1259–1319.
Payne, Rodger A. 2001. “Persuasion, Frames and Norm Construction.”
European Journal of International Relations. 7, 37–61.
Pearlman, Wendy. 2009. “Spoiling Inside and Out: Internal Political Contestation
and the Middle East Peace Process.” International Security. 33 (3), 79–109.
Perla, Hector, Jr. 2005. “Revolutionary Deterrence: The Sandinista
Response to Reagan’s Coercive Policy Against Nicaragua, Lessons
Toward a Theory of Asymmetric Conflict.” Ph.D. dissertation, University of California, Los Angeles.
2008. “Grassroots Mobilization Against US Military Intervention in El
Salvador.” Socialism and Democracy. 22 (3), 143–159.
2009. “Heirs of Sandino: The Nicaraguan Revolution and the US-Nicaragua
Solidarity Movement.” Latin American Perspectives. 36 (6), 80–100.
Bibliography
283
Pichardo, Nelson A. 1988. “Resource Mobilization: An Analysis of Conflicting Theoretical Variations.” Sociological Quarterly. 29 (1), 97–110.
Pickvance, Christopher J. 1999. “Democratization and the Decline of Social
Movements: The Effects of Regime Change on Collective Action in
Eastern Europe, Southern Europe, and Latin America.” Sociology.
33 (2), 353–372.
Piven, Frances Fox & Richard A. Cloward. 1979. Poor People’s Movements:
Why They Succeed, How They Fail. New York: Vintage Books.
Policzer, Pablo. 2006. “Human Rights Beyond the State.” Journal of Human
Rights. 5 (2), 215–233.
Politkovskaya, Anna. 2001. A Dirty War: A Russian Reporter in Chechnya.
London: The Harvill Press.
2003. A Small Corner of Hell: Dispatches from Chechnya. University of
Chicago Press.
Popkin, Samuel. 1988. “Political Entrepreneurs and Peasant Movements in
Vietnam.” In Michael Taylor (ed.), Rationality and Revolution. New
York: Cambridge University Press, 9–62.
Powell, Bill. 1996. “A Religious War?” Newsweek. October 14.
Power, Samantha. 2004. “Dying in Darfur.” The New Yorker. August 30.
Press, Robert, M. 1989. “Sudan Rebels Court US Congressmen.” Christian
Science Monitor. April 3.
Price, Richard. 1998. “Reversing the Gunsights: Transnational Civil Society
Targets Land Mines.” International Organization. 52 (3), 613–644.
Prunier, Gerard. 1998. “The Rwandan Patriotic Front.” In Christopher
Clapham (ed.), African Guerillas. Oxford: James Currey, 119–133.
2004. “Rebel Movements and Proxy Warfare: Uganda, Sudan and the
Congo (1986–1999).” African Affairs. 103, 359–383.
Purdeková, Andrea. 2008. “Repatriation and Reconciliation in Divided
Societies: The Case of Rwanda’s ‘Ingando’.” Refugee Studies Centre
Working Papers.
Quinn-Judge, Paul. 2004. “Russia’s Most Wanted.” Time. October 17.
Rabushka, Alvin & Kenneth Shepsle. 1972. Politics in Plural Societies:
A Theory in Democratic Instability. Columbus, Ohio: Charles E. Merrill.
Radnitz, Scott. 2006. “Look Who’s Talking! Islamic Discourse in the
Chechen Wars.” Nationalities Papers. 34 (2), 237–256.
Rapoport, David C. 2003. “The Four Waves of Rebel Terror and September
11.” In Charles W. Kegley, Jr. (ed.), The New Global Terrorism:
Characteristics, Causes, Controls. Upper Saddle River: Prentice Hall,
36–59.
Rasanayagam, Angelo. 2003. Afghanistan: A Modern History. London: I.B.
Tauris.
284
Bibliography
Rashid, Ahmed. 2008. Descent into Chaos: How the War against Islamic
Extremism is being Lost in Pakistan, Afghanistan and Central Asia.
London: Allen Lane.
Rastegar, Farshad. 1991. “Education and Revolutionary Political Mobilization: Schooling Versus Uprootedness as Determinants of Islamic Political Activism Among Afghan Refugee Students in Pakistan.” Ph.D.
dissertation, University of California, Los Angeles.
Rein, Martin & Donald A. Schoen. 1996. “Frame-Critical Policy Analysis and
Frame-Reflective Policy Practice.” Knowledge and Policy. 9, 85–104.
Reno, William 2009. “The Evolution of Warfare in Africa.” Afrika Focus.
22, 7–19.
Republic of Rwanda. National Unity and Reconciliation Commission
(NURC). 2008. “The Causes of Violence after the 1994 Genocide in
Rwanda.” Kigali, Rwanda: Premier Consulting Group.
Reuter, John. 2004. “Chechnya’s Suicide Bombers: Desperate, Devout, or
Deceived?” Report presented by the American Committee for Peace in
Chechnya and the Jamestown Foundation.
Risse, Thomas. 2000. “‘Let’s Argue!’: Communicative Action in World
Politics.” International Organization. 54 (1), 1–40.
2002. “Transnational Actors and World Politics.” In Walter Carlsnaes,
Thomas Risse, & Beth Simmons (eds.), Handbook of International
Relations. London: Sage Publications, 255–274.
Risse, Thomas, Stephen Ropp, & Kathryn Sikkink (eds.). 1999a. The Power
of Human Rights: International Norms and Domestic Change.
Cambridge University Press.
Risse, Thomas & Kathryn Sikkink. 1999b. “The Socialization of International
Human Rights Norms into Domestic Practices. Introduction.” In Thomas
Risse, Stephen Ropp, & Kathryn Sikkink (eds.), The Power of Human
Rights: International Rights and Domestic Change. Cambridge University
Press, 1–38.
Risse-Kappen, Thomas (ed.). 1995a. Bringing Transnational Relations
Back In: Non-State Actors, Domestic Structures and International Institutions. Cambridge University Press.
1995b. “Bringing Transnational Relations Back In: Introduction.” In
Thomas Risse-Kappen (ed.), Bringing Transnational Relations Back
In: Non-State Actors, Domestic Structures and International Institutions, Cambridge University Press, 3–33.
Rizvi, Gowher. 1990. “The Afghan Refugees: Hostages in the Struggle for
Power.” Journal of Refugee Studies. 3 (3), 144–161.
Rolandsen, ystein, H. 2005. Guerilla Government: Political Changes in
Southern Sudan during the 1990s. Uppsala: Nordic African Institute.
Bibliography
285
Romano, David. 2006. The Kurdish Nationalist Movement: Opportunity,
Mobilization and Identity. Cambridge University Press.
Rotberg, Robert I. (ed.). 2004. When States Fail: Causes and Consequences.
Princeton University Press.
Roy, Olivier. 1986. Islam and Resistance in Afghanistan. Cambridge
University Press.
1989. “Afghanistan: War as a Factor of Entry into Politics.” Central Asian
Survey. 8 (4), 43–62.
1995. Afghanistan: From Holy War to Civil War. Princeton, NJ: Darwin
Press.
Rubin, Barnett R. 1991. “Afghanistan: Political Exiles in Search of a State.”
In Y. Shain (ed.), Governments-in-Exile in Contemporary World Politics. New York: Routledge, 69–91.
2002. The Fragmentation of Afghanistan. New Haven, Conn.: Yale
University Press.
Ruggie, John. 1998. “What Makes the World Hang Together? NeoUtilitarianism and the Social Constructivist Challenge.” International
Organization. 52, 855–885.
Rushton, Simon. 2008. “The UN Secretary-General and Norm Entrepreneurship: Boutros Boutros-Ghali and Democracy Promotion.” Global
Governance. 14, 95–110.
Sageman, Marc. 2004. Understanding Terror Networks. University of
Pennsylvania Press.
Saideman, Stephen, 2001. The Ties that Divide: Ethnic Politics, Foreign
Policy, and International Politics. New York: Columbia University
Press.
2002. “Discrimination in International Relations: Analyzing External
Support for Ethnic Groups.” Journal of Peace Research. 39 (1), 27–50.
Salehyan, Idean. 2007. “Transnational Rebels: Neighboring States as
Sanctuary for Rebel Groups.” World Politics. 59 (2), 217–242.
2008. “No Shelter Here: Rebel Sanctuaries and International Conflict.”
Journal of Politics. 70 (1), 54–66.
2009. Rebels without Borders: Transnational Insurgencies in World
Politics. Ithaca, NY: Cornell University Press.
2010. “The Delegation of War to Rebel Organizations.” The Journal of
Conflict Resolution. 54 (3), 493–515.
Salehyan, Idean & Kristian Skrede Gleditsch. 2006. “Refugees and the
Spread of Civil War.” International Organization. 60, 335–366.
Salehyan, Idean, Kristian Skrede Gleditsch, & David E. Cunningham. 2011.
“Explaining External Support for Insurgent Groups.” International
Organization. 65, 709–744.
286
Bibliography
Salmon, Wesley. 1990. Four Decades of Scientific Explanation. University of
Minnesota Press.
Sambanis, Nicholas. 2004. “Using Case Studies to Expand Economic
Models of Civil War.” Perspectives on Politics. 2 (2), 259–280.
Saurugger, Sabine. 2010. “The Social Construction of the Participatory
Turn: The Emergence of a Norm in the European Union.” European
Journal of Political Research. 49, 471–495.
Sayan, Giyas & Rüdiger Lötzer. 1998. Kurden in Berlin. Berlin: Kurdische
Gemeinde zu Berlin e.V.
Schatz, Edward (ed.). 2009. Political Ethnography: What Immersion Contributes to the Study of Power. University of Chicago Press.
Schimmelfennig, Frank. 2005. “Strategic Calculation and International
Socialization: Membership Incentives, Party Constellations and Sustained Compliance in Central and Eastern Europe.” International
Organization. 59 (4), 827–860.
Schmitz, Hans Peter. 2004. “Transnational NGOs as Agents of Change: Towards
Understanding Their Governance, Leadership and Effectiveness.” Washington, DC: National Science Foundation, Grant SES-0527679 (2005–2008).
2006. Transnational Mobilization and Domestic Regime Change: Africa
in Comparative Perspective. New York: Palgrave Macmillan.
Schöch, Rüdiger. 2008. “Afghan Refugees in Pakistan during the 1980s:
Cold War Politics and Registration Practice.” New Issues in Refugee
Research. Geneva: UN High Commissioner for Refugees.
Schomerus, Mareike. 2007. “The Lord’s Resistance Army in Sudan:
A History and Overview.” HSBA Working Papers. Geneva: Small Arms
Survey/Graduate Institute of International Studies.
2008. “International Involvement for Peacemaking in Northern Uganda.”
In Aaron Griffiths & Catherine Barnes (eds.), Accord. Powers of Persuasion. Incentives, Sanctions, and Conditionality in Peacemaking.
London: Conciliation Resources, 92–96.
Schomerus, Mareike & Kennedy Tumutegyereize. 2009. After Operation
Lightning Thunder. Protecting Communities and Building Peace.
London: Conciliation Resources.
Schroen, Gary C. 2005. First In: An Insider’s Account of How the CIA Spearheaded the War on Terror in Afghanistan. New York: Ballantine Books.
Shahrani, M. Nazif. 1995. “Afghanistan’s Muhajirin (Muslim ‘Refugee Warriors’): Politics of Mistrust and Distrust of Politics.” In E. Valentine
Daniel & John Chr. Knudsen (eds.), Mistrusting Refugees. Berkeley:
University of California Press, 187–206.
Shain, Yossi. 1999. Marketing the American Creed Abroad: Diasporas in the
US and Their Homelands. New York: Cambridge University Press.
Bibliography
287
2002. “The Role of Diasporas in Conflict Perpetuation and Resolution.”
SAIS Review. 22 (2), 115–144.
Shain, Yossi & Ravinatha P. Aryasinha. 2006. “Spoilers or Catalysts?
The Role of Diasporas in Peace Processes.” In Edward Newman &
Oliver Richmond (eds.), Challenges to Peacebuilding: Managing
Spoilers During Conflict Resolution. Tokyo: United Nations University
Press.
Shain, Yossi & Aharon Barth. 2003. “Diasporas and International Relations
Theory.” International Organization. 57 (3).
Shea, Nina. 1998. “A War on Religion.” Wall Street Journal. July 31.
Sheehy, Ann. 1991. “Power Struggle in Checheno-Ingushetia.” RFE-RL
Report on the USSR. November 15.
Sheffer, Gabriel. 2003. Diaspora Politics: At Home Abroad. Cambridge
University Press.
Shlapentokh, Dmitry. 2008. “The Rise of the Chechen Emirate?” Middle
East Quarterly. 15 (3), 49–56.
Sil, Rudra & Peter Katzenstein. 2010a. “Analytic Eclecticism in the Study of
World Politics: Reconfiguring Problems and Mechanisms across
Research Traditions.” Perspectives on Politics 8 (2): 411–431.
2010b. Beyond Paradigms: Analytical Eclecticism in the Study of World
Politics. New York: Palgrave Macmillan.
Simmons, Beth, Frank Dobbin, & Geoffrey Garrett. 2006. “Introduction:
The International Diffusion of Liberalism.” International Organization.
60 (4): 781–810.
Singer, David J. 1961. “The Level-of-Analysis Problem in International
Relations.” World Politics. 14 (1), 77–92.
Singer, Peter W. 2004. “Talk is Cheap: Getting Serious About Preventing
Child Soldiers.” Cornell International Law Journal. 37 (3), 561–586.
2006. Children at War. Berkeley: University of California Press.
Smirnov, Andrei. 2007a. “Umarov Trying to Increase Financial Support
from the Middle East.” Chechnya Weekly. October 25.
2007a. “Will the Rebels Declare the ‘Caucasian Emirate’?” Chechnya
Weekly. November 1.
2007b. “The Caucasian Emirate: A Not so New Idea.” Chechnya Weekly.
November 8.
2007c. “Astemirov Takes Credit for Idea of Caucasian Emirate.” Chechnya
Weekly. November 29.
2007d. “Is the Caucasian Emirate Threat a Threat to the Western
World?” Chechnya Weekly. December 6.
Smith, Christian. 1996. Resisting Reagan: the US Central America Peace
Movement. University of Chicago Press.
288
Bibliography
Smith, Hazel & Paul Stares (eds.). 2007. Diasporas in Conflict: PeaceMakers or Peace-Wreckers? Tokyo: United Nations University Press.
Smith, Tony. 2001. Foreign Attachments: The Power of Ethnic Groups in
the Making of American Foreign Policy. Cambridge, MA: Harvard
University Press.
Snow, David A. & Robert D. Benford. 1992. “Master Frames and Cycles of
Protest.” In Aldon Norris & Carol M. Mueller (eds.), Frontiers in Social
Movement Theory. New Haven, Conn.: Yale University Press.
133–155.
Snow, David, E. Burke Rochford, Jr., Steven K. Worden, & Robert D.
Benford. 1986. “Frame Alignment Processes, Micromobilization and
Movement Participation.” American Sociological Review. 51, 464–481.
Sokirianskaia, Ekaterina. 2010. “Governing Fragmented Societies:
State-Building and Political Integration in Chechnya and Ingushetia
(1991–2009).” Ph.D. dissertation. Department of Political Science,
Central European University, Budapest.
Speckhard, Anne & Khapta Ahkmedova. 2006. “The Making of a Martyr:
Chechen Suicide Terrorism.” Studies in Conflict and Terrorism. 29 (5),
429–492.
Staggenborg, Suzanne. 2008. “Seeing Mechanisms in Action.” Qualitative
Sociology. 31 (4), 341–344.
Staniland, Paul. 2005. “Defeating Transnational Insurgencies: The
Best Offence is a Good Fence.” The Washington Quarterly. 29 (1), 21–40.
Stark, Rodney & William Sims Bainbridge. 1980. “Networks of Faith:
Interpersonal Bonds and Recruitment to Cults and Sects.” The American Journal of Sociology. 85 (6), 1376–1395.
Stedman, Stephen John & Fred Tanner (eds.). 2003. Refugee Manipulation:
War Politics and the Abuse of Human Suffering. Washington, DC:
Brookings Institution Press.
Stein, Gottfried. 1994. Endkampf um Kurdistan? Die PKK, die Tuerkei und
Deutschland. Munich: Bonn Aktuell.
Stenersen, Anne. 2009. “Are the Afghan Taliban Involved in International
Terrorism?” CTC Sentinel. 2 (9), 1–5.
Suhrke, Astri & Frank Klink. 1987. “Contrasting Patterns of Asian Refugee
Movements: The Vietnamese and Afghan Syndromes.” In James T.
Fawcett & Benjamin V. Carino (eds.), Pacific Bridges: The New
Immigration from Asia and the Pacific Islands. New York: Center for
Migration Studies, 85–102.
Sundberg, Ralph, Mathilda Lindgren, & Ausra Padskocimaite. 2010. UCDP
GED Codebook Version 1.0–2011. Department of Peace and Conflict
Research, Uppsala University.
Bibliography
289
Symposium. 2006. “Diffusion of Liberalism.” International Organization.
60 (4), 781–934.
2007a. “Symposium: Multi-Method Work – Dispatches from the Front
Lines.” Qualitative Methods: Newsletter of the American Political Science
Association Organized Section on Qualitative Methods. 5 (1), 9–27.
2007b. “John Gerring, Case Study Research: Principles and Practices
(Cambridge, 2007).” Qualitative Methods: Newsletter of the American
Political Science Association Organized Section for Qualitative and
Multi-Method Research. 5 (2), 2–15.
2008. “McAdam, Tarrow and Tilly’s ‘Measuring Mechanisms of Contention’.” Qualitative Sociology. 31 (4) 333–354.
Tarrow, Sidney. 1998. Power in Movement: Social Movements and Contentious Politics, 2nd edition. New York: Cambridge University Press.
2001. “Transnational Politics: Contention and Institutions in International Politics.” Annual Review of Political Science. 4, 1–20.
2005. The New Transnational Activism. New York: Cambridge University
Press.
2007. “Inside Insurgencies: Politics and Violence in an Age of Civil War”
(Book Review Essay). Perspectives on Politics. 5 (3), 587–600.
2010. “Dynamics of Diffusion: Mechanisms, Institutions, and Scale Shift.”
In Rebecca Givan, Kenneth Roberts, & Sarah Soule (eds.), The Diffusion of Social Movements: Actors, Mechanisms, and Political Effects.
Cambridge University Press.
Tarrow, Sidney & Donnatella della Porta. 2004. “Conclusion: ‘Globalization’, Complex Internationalism, and Transnational Contention.” In
Sidney Tarrow & Donnatella della Porta (eds.), Transnational Protest
and Global Activism. Lanham, MD: Rowman and Littlefield.
Terry, Fiona. 2002. Condemned to Repeat: The Paradox of Humanitarian
Action. Ithaca, NY: Cornell University Press.
Thomson, Susan. 2011. “Reeducation for Reconciliation: Participant Observations on the Ingando.” In Scott Straus & Lars Waldorf (eds.), Remaking Rwanda: State Building and Human Rights after Mass Violence.
Madison: University of Wisconsin Press, 331–339.
Tilly, Charles. 1978. From Mobilization to Revolution. Massachusetts:
Addison-Wesley.
Tishkov, Valery. 2004. Chechnya: Life in a War-Torn Society. Berkeley:
University of California Press.
Toft, Monica Duffy. 2007. “Getting Religion? The Puzzling Case of Islam
and Civil War.” International Security. 31 (4), 97–131.
Tuathail, Gearóid Ó. 2009. “Placing Blame: Making Sense of Beslan.”
Political Geography. 28 (1), 4–15.
290
Bibliography
Tucker, R. W., C. B. Keely, & L.Wrigley (eds.). 1990. Immigration and US
Foreign Policy. Boulder, CO: Westview Press.
Tumelty, Paul. 2006. “The Rise and Fall of Foreign Fighters in Chechnya.”
Terrorism Monitor. January 26.
Turkish Democracy Foundation. 1997. Fact Book on PKK Terrorism.
Ankara.
UN. 1993. “Situation of Human Rights in the Sudan.” Note by the SecretaryGeneral. http://daccess-dds-ny.un.org/doc/UNDOC/GEN/N94/408/18/
pdf/N9440818.pdf?OpenElement. (Accessed May 25, 2010)
UN General Assembly. 1997. The Rights of the Child. 51st Session. Resolution 51/77. http://daccess-dds-ny.un.org/doc/UNDOC/GEN/N97/
768/37/PDF/N9776837.pdf?OpenElement. (Accessed May 20, 2010)
UNHCR. 2001. The State of the World’s Refugees: Fifty Years of Humanitarian Action. Oxford University Press.
2002. “External Evaluation of the UNHCR Shelter Program in Rwanda
1994–1999.” RLSS Mission Report 2000/3. Geneva, Switzerland:
UNHCR.
2003. “UNHCR Assisted Voluntary Repatriation to Afghanistan – Report
from Daily Encashment Centres.” Kabul: United Nations High Commissioner for Refugees.
2010a. “2009 Global Trends: Refugees, Asylum-seekers, Returnees,
Internally Displaced and Stateless Persons.” Geneva: United Nations
High Commissioner for Refugees.
2010b. “Democratic Republic of the Congo, 1993–2003: Report of
the Mapping Exercise on Human Rights Violations.” New York:
UNHCR.
UNICEF. 2000. “Sudan Rebels give UNICEF a guarantee on child soldiers”,
UNICEF press release. www.unicef.org/newsline/00prslpa.htm.
(Accessed July 17, 2010)
2001a. “UNICEF airlifts over 2,500 demobilized child soldiers out of
Sudan combat zone”, UNICEF Press Release. www.unicef.org/newsline/01pr22.htm. (Accessed July 17, 2010)
2001b. “UNICEF Returns Former Child Soldiers Home”, UNICEF Press
Release. www.unicef.org/newsline/01pr71.htm. (Accessed July 17, 2010)
UN Security Council. 1999. “On the Children and Armed Conflict.” Resolution
1261. http://daccess-dds-ny.un.org/doc/UNDOC/GEN/N99/248/59/PDF/
N9924859.pdf?OpenElement.
2000. “On the Children and Armed Conflict.” Resolution 1314. http://
daccess-dds-ny.un.org/doc/UNDOC/GEN/N00/604/03/PDF/N0060403.
pdf?OpenElement. (Accessed May 20, 2010)
Bibliography
291
2001. “On Children and Armed Conflict.” Resolution 1379. http://daccess-dds-ny.un.org/doc/UNDOC/GEN/N01/651/10/PDF/N0165110.
pdf?OpenElement. (Accessed May 20, 2010)
2003. “Children and Armed Conflict.” Resolution 1460. http://daccessdds-ny.un.org/doc/UNDOC/GEN/N03/231/16/PDF/N0323116.pdf?
OpenElement. (Accessed May 20, 2010)
2004. “Children and Armed Conflict.” Resolution 1539. http://daccessdds-ny.un.org/doc/UNDOC/GEN/N04/318/63/PDF/N0431863.pdf?
OpenElement. (Accessed May 20, 2010)
2005. “Children and Armed Conflict.” Resolution 1612. http://daccessdds-ny.un.org/doc/UNDOC/GEN/N05/439/59/PDF/N0543959.pdf?
OpenElement. (Accessed May 20, 2010)
United States Department of State, Office for the Coordinator for Counterterrorism. 2010. “Designation of Caucasus Emirates Leader Doku
Umarov.” June 23. www.state.gov/s/ct/rls/other/des/143564.htm.
(Accessed July 5, 2010)
Uslaner, Eric M. 1991. “A Tower of Babel on Foreign Policy?” In Allan J.
Cigler & A.Burdett (eds.), Loomis Interest Group Politics. Washington,
DC: Congressional Quarterly Inc.
US Senate. 2000. Committee on Foreign Relations. US Commission on
International Religious Freedom: Findings on Russia, China and
Sudan; and Religious Persecutions in the World. 106th Cong., 2d sess.,
May 16.
Ustinova, Mara. 2004. The Chechen Conflict in the Eyes of Prominent
Chechen Intellectuals: An Oral History Approach. Moscow: Institute
for Ethnology and Anthropology, Russian Academy of Sciences.
Van Acker, Frank. 2004. “Uganda and the Lord’s Resistance Army: The
New Order no one Ordered.” African Affairs. 103, 335–357.
van Bijlert, Martine. 2009. “Unruly Commanders and Violent Power
Struggles: Taliban Networks in Uruzgan.” In Antonio Giustozzi (ed.),
Decoding the New Taliban: Insights from the Afghan Field, New York:
Columbia University Press, 155–178.
van der Meeren, Rachel. 1996. “Three Decades in Exile: Rwandan Refugees
1960–1990.” Journal of Refugee Studies. 9(3), 252–267.
Van Evera, Stephen. 1997. Guide to Methods for Students of Political
Science. Ithaca, NY: Cornell University Press.
Van Hear, Nicholas. 1998. New Diasporas: The Mass Exodus, Dispersal
and Regrouping of Migrant Communities. London: UCL Press.
2002. “Sustaining Societies Under Strain: Remittances as a form of Transnational Exchange in Sri Lanka and Ghana.” In Nadje al-Ali & Khalid
292
Bibliography
Koser (eds.), New Approaches to Migration? Transnational Communities
and the Transformation of Home. London: Routledge, 202–223.
van Linschoten, Alex Strick, & Felix Kuehn. 2011. “Separating the Taliban
from Al-Qaeda: The Core of Success in Afghanistan.” CIC Study.
Center on International Cooperation, New York University.
Vatchagaev, Mairbek. 2007. “The Dagestani Jamaat (Part 1).” North Caucasus Analysis. December 13.
Vidino, Lorenzo. 2005. “How Chechnya Became a Breeding Ground for
Terror.” Middle East Quarterly. 7 (3), 57–66.
Vinci, Anthony. 2007. “Existential Motivations in the Lord’s Resistance
Army’s Continuing Conflict.” Studies in Conflict & Terrorism. 30,
337–352.
Vinjamuri, Leslie. 2010. “Deterrence, Democracy, and the Pursuit of International Justice.” Ethics & International Affairs. 24, 191–211.
Waislamah News Network. 2002. The Life and Times of Khattab. www.
maktabah.net/store/Products/ViewProductDetails.aspx?ProductID 29048
&ID 0. (Accessed September 29, 2009)
Waldorf, Lars. 2009. “Transitional Justice and DDR: The Case of
Rwanda.” New York: International Center for Transitional Justice.
Walhbeck, Östen. 1999. Kurdish Diasporas: A Comparative Study. London:
MacMillan.
Walter, Barbara F. 2003. “The Critical Barrier to Civil War Settlement.”
International Organization. 51 (3), 335–364.
Waltz, Kenneth. 1979. Theory of International Politics. Reading, MA:
Addison-Wesley.
Ware, Robert Bruce. 2005. “A Multitude of Evils: Mythology and Political
Failure in Chechnya.” In Richard Sakwa (ed.), Chechnya: From Past to
Future. London: Anthem Press, 79–116.
Watts, Nicole, F. 2004. “Institutionalizing Virtual Kurdistan West: ProKurdish Politics in Western Europe.” In Joel Migdal (ed.), Boundaries
and Belonging: States and Societies in the Struggle to Shape Identities
and Local Practice. Cambridge University Press,
Weidmann, Nils B. 2007. “From Sparks to Prairie Fires: Spatial Mechanisms
of Group Mobilization.” Paper presented at the Annual Convention of
the American Political Science Association, Chicago.
Weil, Martin. 1974. “Can the Blacks Do for Africa what the Jews Did for
Israel?” Foreign Policy. 15, 109–130.
Weiner, Myron. 1992. “Security, Stability and International Migration.”
International Security. 17(3), 91–126.
1996. “Bad Neighbors, Bad Neighborhoods.” International Security.
21 (1), 5–42.
Bibliography
293
Weinstein, Jeremy. 2007. Inside Rebellion: The Politics of Insurgent Violence. New York: Cambridge University Press.
Wendt, Alexander. 1992. “Anarchy is What States Make of It: The Social
Construction of Power Politics.” International Organization. 46 (2),
391–425.
1999. Social Theory of International Politics. Cambridge University Press.
Weyland, Kurt. 2007. Bounded Rationality and Policy Diffusion: Social
Sector Reform in Latin America. Princeton University Press.
2009. “The Diffusion of Revolution: ‘1848’ in Europe and Latin
America.” International Organization. 63, 391–423.
White, Paul J. 2001. Primitive Rebels or Revolutionary Modernizers? The
Kurdish National Movement in Turkey. London: Zed Books.
Wight, Colin. 2002. “Philosophy of Science and International Relations.” In
Walter Carlsnaes, Thomas Risse, & Beth Simmons (eds.), Handbook of
International Relations. London: Sage Publications, 23–51.
2006. Agents, Structures and International Relations: Politics as Ontology. Cambridge University Press.
2007. “A Manifesto for Scientific Realism in IR: Assuming the
Can-Opener Won’t Work!” Millenium: Journal of International Studies.
35 (2), 379–398.
Wilhelmsen, Julie. 2004. When Separatists Become Islamists: The Case of
Chechnya. Kjeller, Norway: The Norwegian Defence Research
Establishment.
Williams, Brian Glyn. 2000. “Commemorating ‘The Deportation’: The Role
of Memorialization and Collective Memory in the 1994–96 Chechen
War.” Memory and History. 12 (1), 101–134.
2003. “Shattering the Al-Qaeda-Chechen Myth (Part II).” Chechnya
Weekly. November 6.
2005a. “The ‘Chechen Arabs’: An Introduction to the Real Al-Qaeda
Terrorists from Chechnya.” Terrorism Monitor. May 5.
2005b. “Turkish Volunteers in Chechnya.” Terrorism Monitor. April 6.
2007. “Allah’s Foot Soldiers An Assessment of the Role of Foreign
Fighters and Al-Qa‘ida in the Chechen Insurgency.” In Moshe Gammer
(ed.), Ethno-Nationalism, Islam and the State in the Caucasus: PostSoviet Disorder. London: Routledge, 156–178.
Williams, Paul D. 2007. “From Non-Intervention to Non-Indifference: The
Origins and Development of the African Union’s Security Culture.”
African Affairs. 106, 253–279.
Wolf, Frank, R. 2001. “Op-Ed: Peace In Sudan Needs To Be International
Priority.” http://wolf.house.gov/index.cfm?sectionid 34&sectiontree 6,34
&itemid 483. (Accessed July 20, 2010)
294
Bibliography
Wood, Elisabeth Jean. 2003. Insurgent Collective Action and Civil War in El
Salvador. New York: Cambridge University Press.
2008. “The Social Processes of Civil War: The Wartime Transformation
of Social Networks.” Annual Review of Political Science. 11, 539–561.
2009. “Armed Groups and Sexual Violence: When is Wartime Rape
Rare?” Politics and Society. 37 (1), 131–162.
2010. “Sexual Violence During War: Variation and Accountability.” In
Alette Smeulers & Elies van Sliedregt (eds.), Collective Crimes
and International Criminal Justice: An Interdisciplinary Approach.
Antwerp: Intersentia, 295–322.
Woodward, Peter. 2006. US Foreign Policy and the Horn of Africa. Burlington, VT: Ashgate.
Wucherpfennig, Julian, Nils Metternich, Lars-Erik Cederman, & Kristian
Skrede Gleditsch. 2012. “Ethnicity, the State, and the Duration of Civil
War.” World Politics. 64 (1), 79–115.
Yousaf, Mohammad & Mark Adkin. 1992. The Bear Trap: Afghanistan’s
Untold Story. Lahore: Jang.
Zahab, Mariam Abou & Olivier Roy. 2004. Islamic Networks: The
Pakistan-Afghan Connection. London: C. Hurst.
Zolberg, Aristide, Astri Suhrke, & Sergio Aguayo. 1986. “International
Factors in the Formation of Refugee Movements.” International Migration Review. 20 (2), 151–169.
1989. Escape from Violence: Conflict and the Refugee Crisis in the
Developing World. New York: Oxford University Press.
Zürcher, Christoph. 2007. The Post-Soviet Wars: Rebellion, Ethnic Conflict,
and Nationhood in the Caucasus. New York University Press.
Index
Abdelal, R., 173, 179
abductions, 120, 123 4, 126, 128,
134 9, 141 2, 144, 147, 251, 258
See also hostage taking; kidnapping
Abdulkhadzhie, Aslambek, 55
Abkhazia, 46
Abu Walid al Ghamdi, 50, 57
accountability, 65, 131, 135, 145, 165
“accountability and reconciliation”,
145
Acharya, Amitaev, 195, 236
Acholi, 124 5, 127 30, 133, 136 7,
140, 144, 147
diaspora, 123, 138 9, 249, 252
See also Lord’s Resistance Army;
Uganda
Achverina, Vera, 153
Adamson, Fiona, 123, 200, 217,
219 20, 223, 225, 244 5, 250 2,
255 6
adaptation, social, 14 15, 122, 218 19,
255
and appropriateness, 179
and entrepreneurship, 173 201, 220,
257
and Sudan, 147
defined, 174, 179
vs. emulation, 179
Afghanistan
Hezb e Islami, 99, 101 3
Jamiat e Islami, 99, 103 4
Junbesh e Milli, 101
Kabul, 101, 103, 106, 254
Kandahar, 104
Northern Alliance, 104
Operation Enduring Freedom, 106
People’s Democratic Party of
Afghanistan, 97, 101, 103
African Union, 180
agency, 4, 16, 23
collective, 68
of refugees, 119, 249
political, 89, 245
vs. structure, 178
agenda setting, 84
agent based modeling, 11, 15, 17, 20,
22, 174 5, 178, 181 4, 188, 195,
217, 220, 252, 254 5, 257 8
Aguayo, Sergio, 96
aid, 106 7, 126, 141, 161, 225, 250
emergency, 119
food, 108
humanitarian, 60, 98, 110, 113, 139,
161, 164
military, 171
refugee, 98
See also NGOs
Al Jazeera, 56
al Qaeda, 65, 106, 231
See also terrorism
Algeria, 86
alienation, 38, 193, 220
allegiance/loyalty, 25, 94, 102 3,
116 17, 119, 252
alliances, 33, 177, 222, 255
and security dilemma, 222
burden sharing, 222
transnational mobilization of, 256 7
Amin, Idi, 110, 127 8, 136
amnesty, 124, 130 2, 142 4, 146, 217
Amnesty International, 126, 167
analytic eclecticism, 11, 21, 212 13
Andvig, Jens Christopher, 153
appropriateness, social, 38 9, 174, 178,
180, 185, 192 3, 215
“logic of appropriateness”, 214
and free riding, 214
and norm entrepreneurs, 180 1
and scientific realism, 209
and social adaptation, 179, 185
295
296
appropriateness, social (cont.)
See also norms
Arab fighters, 43, 52, 57 8
Arusha Peace Agreement, 112, 117 18
ASEAN, 180
assassination, 38, 219
of Khattab, 59
assimilation, 77, 80
assumptions, 62, 66, 177, 192, 209, 229
“as if” assumptions, 209
about collective actors, 68
about diffusion, 174
about entrepreneurship, 199
about impact of return, 90
about ontological reality, 207
about rationality, 93
about resources, 39
about Rwandan genocide, 114
about social identity, 177
about triangulation, 23
about Ugandan conflict, 128
demand side, 153
disaggregation of, 12
homogenizing, 236
realist, 11
Astemirov, Anzor, 61
asylum, political, 74, 82, 86, 176
and corruption, 83 4
Atatürk, Kemal, 73
Kemalists, 73, 77
atrocities, xiii, 55, 120
maiming/mutilations, 128, 130,
135 9, 147, 249
Auma (Lakwena), Alice, 127 8, 136 7
autonomy, 224, 226 8
and Joseph Kony, 227
Kurdish, 81
Axtell, R., 183
Babitsky, Andrei, 50
Baiev, Khassan, 60
Bainbridge, William Sims, 37
Bakke, Kristin, 14, 21, 69, 91, 123,
136, 199 200, 219 21, 223, 225,
247 9, 252, 254 5
Barfield, Thomas, 105
bargaining, 6, 8, 223
leverage, 70, 223 4
triangular, x, 224, 226
Basayef, Shamil, 45 52, 54 5, 58 9
Index
battlefield dynamics, 154, 157, 170
socialization, 105, 248 9, 254
Bayesian inference, 19, 22, 212, 234
See also rationalism/rational choice
Belgium, 75
and Rwanda, 107
and southern Sudan, 139
Bellamy, Carol, 156, 160
Bennett, Andrew, 15, 17 18, 26, 67,
87, 178
Bhaskar, Roy, 205
bias, 19, 23, 126, 208, 212, 253
and norms, 179, 186
confirmation, 212
ethnic, 130
media, 186
Bigombe, Betty, 129, 144
Black, Richard, 90
Blattman, Christopher, 5
boomerang pattern, 72, 84, 151,
154 8, 161, 171, 245 6
See also norms
Boutros Boutros Ghali, 180
Brady, Henry, 210
brokerage, 5, 35, 37 9, 46, 60, 68,
69, 95, 217 18, 221, 244, 252,
254 5
See also diffusion, mediated
Brown, Elijah, 165
Brownback, Sam, 160, 163 4,
166
Brubaker, Rogers, 177
Bruinessen, Martin van, 76 7
Budennovsk hospital, 48, 53, 58
Buhaug, Halvard, 182 3, 239, 242
Bush, George W., 162 4
camps
IDP, 146
LRA, 124, 130, 132, 136, 139 40,
250
re education, 116, 249
refugee, 92, 98 9, 108 9, 112 13,
120, 153, 159, 254
solidarity, 114
training, 38, 46 7, 49, 56 9, 100,
199, 231, 247 8, 253
Canada, 180
capitalism, 79
Carter Center, 132, 140
Index
causality, 4
causal complexity, 22
inverse, 243
Cederman, Lars Erik, 7, 240, 245
Central African Republic, 120, 124,
138, 146, 226, 249, 251
charities, 44, 59 60, 139
See also NGOs
Chechnya, xiii, 14, 21, 31, 125, 136,
173, 199 200, 217, 219 20,
225 6, 247 8, 251 2, 255 6, 258
Grozny, xiii, 44, 49, 58
Checkel, Jeffrey, 34 5, 63, 66 7, 91 2,
149, 174, 178, 206, 209, 211 12,
216, 235, 253
child soldiers, xiii, 3, 14, 24, 120, 142,
157, 218 21, 225, 245 6, 251,
254
children, 120, 126, 135, 138
and genocide, 115
China, 237
and Uighurs, 192 3
liberty in, 192
Christianity. See Holy Spirit Movement;
missionaries
churches. See under Sudan
citizenship, 214
German, 78, 244
civil society, 85 6, 99, 114
Acholi, 129, 144
and PKK, 244 5, 251
European, 180
Sudanese, 151, 155, 161, 171
US, 246
civilizational rallying, 65
cleavage (defined), 177
Clinton, Bill, 162 4
coalition building, 34, 84, 86, 256
coercion, 13, 16, 116, 182, 234 5
and PKK, 80
and Rwanda, 115
defined, 235
Cohen, Dara Kay, 92
Cold War, 98, 231, 258
post Cold War system, 3, 32, 64,
120 1, 133, 153
Coleman, James, 181
collective action, 39, 70, 181, 214 15, 222
Collier, Paul, 64 5
Colombia, 32, 231
297
colonialism
and Caucasus, 46
Belgian (Rwanda), 107
British (Uganda), 127, 136
communications, 58, 140
communism, 32, 43, 53
comparative politics literature, 3, 26,
66, 229, 232
competition, 234 5, 241, 258
defined, 235
functional, 215
complexity, 113, 135, 186, 201, 220,
222, 229
causal, 22
of theory, 228, 230
concessions/promises, 14, 124, 156 8,
165 6, 246
See also rewards
conflict entrepreneurs.
See entrepreneurs, conflict
Congo, Democratic Republic of, 116,
120, 124, 131, 138 9, 141 2, 226,
249, 251
Garamba National Park, 133, 141,
146
Operational Lightning Thunder, 146
See also Zaire
consequences, logic of, 214 15
constructivism, 7, 10, 12, 18 19, 210,
213 15, 229
and learning, 13
and socialization, 92
positivist, 20
contentious politics, 18, 69, 251 3
radical contention, 38
correlational analysis, 4, 34
See also statistical/regression analysis
corruption, 130
of asylum process, 83 4
cost/benefit calculations, 4, 7 8, 13, 19,
156, 234
and prospective insurgents, 238
counterfactuals, 23, 210, 220
covering laws, 205 8
culture, 78, 215, 237
cultural associations, 75, 176
cultural autonomy, 81
cultural death, 187
cultural liberties, 192
festivals, 78, 85, 114, 244
298
cumulation, 15, 17, 26, 212 21,
229 30, 235
of errors, 23
See also generalizability
Cunningham, David, 242
Cyprus, 173
Dagestan, 43, 47 9, 52 3, 59 60
Dagne, Ted, 165
deductive nomological approaches,
207 8
demobilization (defined), 151 2
as elite driven, 166
democracy, 46, 130, 133 4, 143
and accountability, 131
and LRA manifesto, 130 1
and mobilization, 245, 256
and risk of conflict, 239
democratization, 66, 112
promotion of, 180
Democratic Republic of Congo.
See Congo, Democratic Republic of
demographics, 243
demographic change, 215, 218 19
development, economic, 66, 77, 106, 117
diaspora (as unitary actor), 65 7, 87, 251
diffusion, 35, 253
defined, 12, 35
mediated (defined), 35
relational vs. non relational, 35, 37,
124
disarmament, demobilization, and
reintegration, 91, 116, 145
and United Nations, 180
discrimination, 44, 80, 251
diversity, ethnic. See under ethnicity
Dolan, Chris, 126, 132
donations, 64, 81 4, 244
forced, 82 3
See also extortion
Dostum, Rashid, 101
drought, 231
as metaphor, 233
drugs, 106, 231
See also trafficking
Dudayev, 54
Dudayev, Dzhokhar, 43 6, 51 2, 61
Eccarius Kelly, Vera, 79
efficiency, functional, 178, 215, 217 19
Index
El Salvador, 19, 173, 245
emotion, 13, 19, 25, 79, 184
emulation, 12 13, 35 9, 215, 218, 235,
238, 241, 251, 258
Chechnya, 14, 21, 50, 54, 56 7, 61,
247 8
defined, 179, 234
LRA, 14, 122 4, 136, 138, 147 8, 250
vs. learning, 234
vs. social adaptation, 179, 219
England, 75, 82, 86
colonialism, 128, 136
entrepreneurs, conflict, 70, 244
entrepreneurs, norm (defined), 174
epistemic communities, 33, 62
epistemology, 19 20, 205 6, 229
epistemological opportunism, 11
Epstein, J. M., 183
Eritrea, 170 1
essentialism, 67
Ethiopia, 159, 171
ethnicity, 7, 103, 115, 118, 153, 240
ethnic activists, 180
ethnic diversity, 111, 238, 240
ethnography, 18 20, 22, 59, 73, 121
Europe, 22, 32, 65, 214, 231
and Kurds, 73 88, 244 5, 251 2,
255 6
and revolution, 234
and Uganda, 111
Europe, Eastern, 231
European Union, 74, 180
and Turkey, 74, 85 6, 245
EU institutions, 86
events data, 18, 24, 211 12, 253
evolution, 14
evolutionary selection, 215, 218
unintended, 215
exile, government in, 86, 97,
113
experiments, 210
extortion, 83, 113, 244
extra judicial executions.
See assassination
failed/failing states, xiii, 6, 33
falsification, 207
family reunification, 74
Fathi al Shishani, Ali, 43, 46, 59
Fearon, James, 218
Index
feasibility, 17, 24 5
feudalism, 79
Fikes, Deborah, 165
Finnemore, Martha, 180, 185
Finnström, Sverker, 128, 131,
134
foreign aid. See aid
formal modeling, 19 20, 206,
211
See also game theory; rationalism/
rational choice
Forsberg, E., 183
foundations. See charities
framing, 40
defensive, 125, 134
defined, 69
frame alignment, 70
frame bridging, 219
frame extension, 219
strategic, 68, 71, 244
France, 75, 83
revolution in, 231, 234
free riding, 39, 214
Frist, Bill, 163 4
Fujii, Lee Ann, 109
fundamentalism, 50
fundraising, 83, 87
See also donations
game theory, 184
See also formal theory; rationalism/
rational choice
Garang, John, 163 5, 169
gatekeepers, 51, 247
Gates, Scott, 153
generalizability, 13, 172, 216, 219 21,
230, 232
See also cumulation
genocide, 108, 115, 117
“genocide ideology”, 115
George, Alexander, 17 18, 210
Georgia, 46, 57
Germany, 222
and Turkey, 73 86, 175 6, 188, 192,
200, 217, 219 20, 244, 256
Gerring, John, 16 18, 67, 91, 206, 210
Girardin, Luc, 240, 245
Gleditsch, Kristian Skrede, 6, 182 3,
238 42, 245
Graham, Franklin, 162 3
299
Greece, 32
greed. See under grievances
grievances, 54, 239
and child soldiers, 153
and Uganda, 124, 128, 134, 139,
141 2, 147 8, 195
of Kurds, 86
vs. greed, 64
growth, economic, 106, 239
guerilla warfare, 57
guest workers, 76, 176
Haider, Ziad, 192 3
Hall, Peter, 229
Hamberg, Stephan, 14, 217, 219 20,
223, 225, 245 6, 251, 254
Haq, Abdul, 104
Harpviken, Kristian Berg, 188, 219 20,
223, 225, 227, 248 9, 251, 254
Hedström, Peter, 181
hegemony, 78, 80, 147, 235
hegemonic socialization, 215, 218
state, 78
US, 79
Hempel, Carl, 205, 207
Hill, S., 182
historical institutionalism, 229
HIV/AIDS, 115
Hoffmann, Matthew, 184 9, 191,
196
Holy Spirit Movement, 124, 127 9,
137
homeland groups, 223
homeland politics, 70
hostage taking, 38, 53 4
Beslan school siege, 41, 53 4, 56 7,
60
Budennovsk hospital, 48, 54 5
Kizlyar Pervomaiskoye crisis, 53
See also abductions; kidnapping
humanitarian law, 143
Hume, David, 9, 210
identity options, 79, 200
incentives, selective, 39, 93 4
induction, 16, 211, 221
information
asymmetries, 223
centers, 75, 85
sharing, 158
300
innovation
functional, 218
institutional, 93
policy, 234 5, 255
state to state, 258
tactical, 35, 39, 40 1, 43, 60 2, 215,
251 3
technological, 153
insurgents, transnational (defined), 32
interdependence, 233, 238, 256
and refugees, 119
social construction of, 234
internally displaced persons, 112, 133,
141, 146, 153
See also under camps
International Criminal Court, 120,
122, 124 5, 132 4, 138, 142, 147,
220, 250
See also prosecution
international relations subfield, 4, 6,
195, 205, 213, 229, 232
as competing isms, 213
IR/sociology, 12, 14 15
internet, 35, 53, 123, 125, 131, 253
Kavkaz Center, 42, 52, 55
interpretism, 19, 23
interpretivists, 210
interview methods, 19 20, 22 4, 93 4,
237
Iran
and Afghanistan, 97, 101
and Kurds, 84
Iraq
and Kurds, 75, 84, 226
and suicide terrorism, 57
Islam
mosques, 36, 49, 59, 247, 253
Salafi, 45, 49, 56, 257
sharia law, 44, 46, 49, 170, 225
Shia, 101
Sunni, 100
Islamic International Peacekeeping
Brigade, 47
Italy, 75, 86
Ittehad, 101
Japan, 180, 222
Johnson, Douglas, 158 9
Jordan, 42
Juba peace talks, 132, 134 5, 139, 141 7
Index
Kadyrov, Ramzan, 46
Kagame, Paul, 112
Kaldor, Mary, 64
Kalyvas, Stathis, 177
Karakoram Highway, 192 3
Karzai, Hamid, 97, 104 5
Kashmir, 57
Katzenstein, Peter, 172, 213
Kebedov, Bagaudin, 48, 52, 60
Keck, Margaret, 84, 154 5, 245
Kent, Randolph, 117
Kenya, 86, 109, 165, 231
Khattab, 42 3, 46 51, 54, 56 60,
199 200
kidnapping, 41, 53, 56 7, 247
See also abductions; hostage taking
Kiir, Salva, 160
King, Charles, 52, 212
kinship, 25
as framing strategy, 70, 244
Kony, Joseph, 120 1, 126 9, 131 4,
136 46, 217, 227
Kosovo, 88
Kuhn, Thomas, 205, 207, 213 14, 229
Kuran, T., 180
Kurds
culture, 75, 79, 81, 83, 85, 87, 176,
244, 255
Islamic Justice and Development
Party (AK), 74
Kurdistan Workers’ Party (PKK), 63,
72 88, 123, 175 6, 188, 199, 217,
219 20, 244, 251, 255 6
Kydd, Andrew, 71
Laitin, David, 177, 218
Lakatos, Imre, 205, 207 8, 213 14,
229
Lakwena, Alice. See Auma (Lakwena),
Alice
Langi, 127, 136
language, 71, 241
and LRA, 131
in Rwanda, 114
Kurdish, 74, 78, 85, 87, 244
Tatar, 187
Latin America, 231, 234
learning
and socialization, 108, 118
complex, 182, 235
Index
learning (cont.)
defined, 234
transformative, 90 1, 112, 114 15, 118
Lebanon, 74 6, 84
Leggewie, Claus, 76
legitimacy, 35, 71, 84 5, 178, 209,
213 15, 217 18, 228, 230, 240
“legitimacy” solution, 214
mechanisms, 214, 217
Leiby, Michele, 25
levels of analysis, 209, 214 15
leverage politics, 84
See also persuasion/lobbying
liberalism/neoliberalism, 213, 215, 229
liberalization, 87, 192
liberty, 192
associational, 245, 256
economic, 192
political, 176
religious, 163, 192
Lichbach, Mark, 20
Lischer, Sarah Kenyon, 188, 219 20,
223, 225, 227, 248 9, 251, 254
lobbying. See persuasion/lobbying
localization, norm, 195, 201, 237
defined, 236
looting, 124, 137, 139, 142, 251
Lord’s Resistance Army, 14, 120 48, 217,
219 20, 223, 226 7, 250 2, 254
as irrational, 121, 124 5, 129, 133,
135, 139, 143, 147 8
loyalty. See allegiance/loyalty
Lyon, Alynna, 176
Mahoney, James, 213
Malet, David, 32
Mamdani, Mahmood, 109 11, 113 14
manipulation theory, 210
Mantilla, Giovanni, 180
marginalization, 127, 133, 195, 244
Marxism, 79, 81
Maskhadov, Aslan, 46, 48 9, 51 2,
57, 61
McAdam, Douglas, 17 18, 20
measurement error, 207
mechanisms (defined), 7
Melvin, Neil, 65
memoir literature, 24, 47, 164
memory, collective, 43
Merton, Robert, 9
301
metaphysics, 207 8
methodological individualism, 10, 181
Metternich, Nils, 240
Mexico, 32
Mgbako, Chi, 116
middle range frameworks, 14
migration, 68, 188, 235
and Afghanistan, 99, 104
and Rwanda, 117, 119
as diffusion mechanism, 251, 257
economic, 82, 176
from Turkey, 73, 77, 82, 176
of combatants, 258
Miguel, Edward, 5
millenarianism, 128
mimicking, 179
See also emulation
Min, Brian, 240
Minear, Larry, 117
missionaries, 123, 127 8, 133
mixed method research, 8
mobilization, resource. See resource
mobilization
Mozambique, 154
multinational corporations, 33
Museveni, Yoweri, 110, 125 31, 133,
136, 141 6, 227
Muslim Brotherhood, 97
mutilations. See under atrocities
Nabi, Mohammad, 100
naming and shaming. See shaming
nationalism, 41, 44 5
“long distance”, 65
Chechen, 247
Kosovar, 88
Kurdish, 77, 79, 88, 176
Tatar, 187
Turkish, 73, 79
Uighur, 193
NATO, 88, 245
neoliberal institutionalism, 213
See also liberalism/neoliberalism
neorealism. See under realism
Netherlands, 75, 83
networks
conflict networks, 69, 75 6
social networks, 25, 68, 92, 228
See also transnational advocacy
networks
302
NGOs, 3, 12, 150, 223, 254
Amnesty International, 126, 167
and boomerang model, 72
and Kurds, 86
and LRA, 127, 133, 142, 146
and Rwanda, 114
and Sudan, 157, 220, 225
Caritas, 139
Coalition to Stop the Use of Child
Soldiers, 160, 167
Human Rights Watch, 159, 167
Samaritan’s Purse, 162 3
Save the Children, 143
World Vision, 144
See also transnational advocacy
networks
Nome, Martin Austvoll, 14, 123, 217,
219 20, 252 5
non state actors, 4 5, 33, 64, 150, 258
and definition of “transnational
insurgents”, 32
and drugs, 231
Non State Actor (NSA) dataset, 240
non state armed groups, 149 50, 172
See also NGOs
norm entrepreneurs. See entrepreneurs,
norm
norms
constitutive, 173, 176 8, 184, 187,
192 3, 199
life cycle, 185
Western, 111
novelty, 207 8
use novelty, 208, 212
O’Neill, Bard, 92
Obama, Barack, 122
Obote, Milton, 110, 127 8, 136
Öcalan, Abdullah, 73 4, 82, 86, 130
Okawara, Nobuo, 213
omitted variables, 23, 214, 227
ontology, 10, 21, 207 8, 210, 229
open polity models. See polity (closed
vs. open)
Oppenheim, Paul, 207
opportunity costs, 221, 223
opportunity structures. See political
opportunity structures
Otti, Vincent, 145
outbidding, ethnic, 68, 71, 244
Index
Pakistan, 96 100, 103 7, 119, 231,
248
and Chechnya, 43, 60
and Pashtuns, 225 6, 228
and resource control, 248
and Taliban, 223, 225
and US drones, 231
and Uighurs, 192
Inter Services Intelligence, 100
Palestine, 57
paradigms, 213 14, 229 30
paradigm wars, 205
parsimony, 26, 206, 228, 230
Pashtuns, 103, 105, 225 6, 228
refugees, 225, 228
patronage, 50, 94, 116
financial, 50
peacebuilding, 90
peacekeeping, 95, 116, 162
Islamic International Peacekeeping
Brigade, 47
peacemaking, 162
perception (fuzzy vs. perfect), 189, 197,
199
persuasion/lobbying, 5, 12 13, 22, 68,
71, 72, 88, 154, 158, 180 2, 209,
244
Peru, 25, 237
pluralism, 21
social theoretic, 12
structured, 212, 229 30
Poland, 86
political economy, 4, 64
political opportunity structures, 6, 67,
215
polity (closed vs. open), 5 6, 8 9, 184,
187
positivism, 11, 23, 205, 207 8, 210,
229
positivist constructivism, 20
post communism, 32
post modernism, 206, 209
poverty, 52
and child soldiers, 152 3
power, 16, 71, 126, 178, 213, 228, 230
and identity, 71, 79, 240
and international rhetoric, 147
and logic of consequences, 214
and narratives of conflict, 243
and public goods, 214
Index
power (cont.)
balance of, 34, 60, 159
coercive, 235
concentrated vs. dispersed, 33
effect of exclusion, 240
Ethnic Power Relations database,
240
material, 209, 215, 218
of brokers, 68
of foreign actors, 61, 84, 241
of norm entrepreneurs, 174, 220
of socialization, 109
of TANs, 150
state vs. hinterland, 158
power transition, 12, 215
Power, Samantha, 163
power sharing, 95, 111, 117
pragmatism, 11, 21
prediction, 10, 158, 206 8, 212, 222,
226 7, 238 9, 243
primordialism, 67
process tracing, xiv, 5, 8, 11, 13, 27, 66,
206, 211 12, 222, 227, 232, 253,
255
promises. See concessions/promises
propaganda, 42, 59, 80, 113, 131, 144
property, 94, 119
damage to, 106
laws, 117
property space, 178, 221
repossession of, 117 18
prosecution, 124, 142, 147, 250
See also International Criminal Court
protection money. See extortion
proxy warfare, 123, 125, 138, 250
Prunier, Gerard, 109 10
public goods, 214
Purdeková, Andrea, 113, 115
Putin, Vladimir, 41
quantitative methods. See statistical/
regression analysis
Raduyev, Salman, 55, 60
Ramzan, Amir, 50
rape, 16, 150
Rastegar, Farshad, 99
rationalism/rational choice, 6 7, 10,
12 13, 18 19, 92, 179, 229,
234 5, 241
303
bounded rationality, 94, 234, 246
See also formal theory; game theory
rationalism/rational choice, 44, 170
realism, 9, 11, 229
critical realism, 205
neorealism, 213, 215
scientific realism, 10 11, 21, 206 10,
228
receptivity, 109, 182, 187, 196 7, 252,
257
reconciliation, national, 25, 114 15,
117, 129 30, 143
“accountability and
reconciliation”, 145
reconstruction, post conflict, 64
refugees. See under camps
regression. See statistical/regression
analysis
regularity theory, 210
Reich, Simon, 153
relational diffusion. See under diffusion
religion (as transnational force), 125,
248, 256 7
remittances, 70
repatriation, 93
Afghanistan, 101
Rwanda, 113
residual variance, 8
resource mobilization, 35, 40, 43, 61 2,
68, 71, 88, 93, 122 4, 133 5, 147,
244, 247, 250 3
responsibility to protect, 219
returnee. See repatriation; violence,
post return
revolution, 39, 99, 215, 231, 235, 258
rewards, 39, 94
and human rights behavior, 124,
138, 141
and Islam, 49 50
as selective incentives, 93
See also concessions/promises
Rothchild, D., 182
Roy, Olivier, 100
Rushton, Simon, 180
Russia
and Chechnya, 31, 40 4, 48,
50 61
and Tatarstan, 187
Dubrovka/Nord Ost theater siege,
41, 53, 56 7, 60
304
Russia (cont.)
Russians, 239
See also Soviet Union
Rwanda, 91, 220, 226 7, 248 9, 251,
254, 257
Sadulayev, Abdul Halim, 45
Sagemen, Marc, 65
Salehyan, Idean, 6, 8, 223, 238, 241 3
sanctions, 155 6
and Sudan, 155, 157, 162, 165
sanctuaries, 6, 31, 104, 222
Saudi Arabia, 42, 44, 59 60
Saurugger, Sabine, 180
scapegoating, 110
Schmitz, Hans Peter, 14, 69, 193, 217,
219 20, 223, 225, 227, 249 50, 254
scientific realism. See under realism
scope conditions, 20, 174 5, 183,
216 17, 220, 222, 230, 255
secularism, 51, 124, 131
security dependence, 222
security dilemma, 249
alliance, 222
post return, 95
security entrapment, 90 1, 97, 108,
113, 118 19, 248
selective incentives. See incentives,
selective
self determination, 41, 44 6, 51, 60
shaming, 14, 124, 138, 146 7, 149 52,
155 8, 161, 165, 209, 218, 220,
246, 254
Sierra Leone, 167
Sikkink, Kathryn, 84, 154 5, 180, 185,
245
Sil, Rudra, 213
Singer, Peter, 153
slavery, 132, 158, 162 3
anti slavery NGOs, 161
sex slaves, 120
small group dynamics, 15
Smirnov, Andrei, 60
Snyder, Jack, 172
social adaptation. See adaptation, social
social movements, 32 3, 35 6, 38, 62,
67, 69 71, 122
socialization (defined), 15, 21, 90 1
military, 92
Sokirianskaia, Ekaterina, 47, 49 50, 59
Index
solidarity groups, 85, 92
Somalia, 231
South Africa, 86, 154
apartheid, 231
Soviet Union, 98, 153, 193, 231, 239
See also Russia
Spain, 32, 86
spirituality. See Holy Spirit Movement
Sri Lanka, 231
Staniland, Paul, 34, 36
Stark, Rodney, 37
statistical/regression analysis, 20, 34,
183 4,206,210 11,216,220 1,238
structural adjustment programs, 131
structural theory, 10, 209, 228
structural factors, 19, 214, 217, 221,
228
structuration theory, 214
Sudan
inequality in, 159
New Sudan Council of Churches,
151, 161 6, 172, 246
Operation Iron Fist, 132, 140, 143
Operation Lightning Thunder, 146
SPLA/M, 132, 140, 144, 149 72,
218 20, 223, 225 6, 245 7, 250,
254
Sudan Peace Act, 163 4
Suhrke, Astri, 96
Swedberg, Richard, 181
Sweden, 82
Switzerland, 75, 83
symbolic acts, 69, 78, 85, 166
Syria, 74 6, 86, 244
Damascus, 76, 82 3, 244
tactical innovation. See under
innovation
Tajikistan, 42, 46
Tajiks, 103
Taliban, 97 107, 223, 225, 248 9
Tamils, 231
Tanzania, 107
Arusha, 112, 117 18
Tarrow, Sidney, 5, 17, 18, 20, 252
Tatarstan, 59, 187, 196
technology, 239
information, 219
spying, 228
weapons, 153
Index
terrorism, 31, 65, 258
al Qaeda, 65
Beslan school siege, 41, 53 4, 56 7, 60
Chechen, 41, 53 4, 58, 258
Dubrovka/Nord Ost theater siege,
41, 53, 56 7, 60
September 11, 31, 105, 125
suicide, 53 4, 57 8
See also al Qaeda
Thomas, Craig, 18
Tilly, Charles, 17 18, 20
Tishkov, Valery, 43
torture, 38, 135, 258
See also atrocities
trade, 192, 239
illicit, 106
slave, 132
weapons, 140
trafficking
arms, 3, 39, 59, 124, 140 1
drug, 106
human, 84, 132
training. See under camps
transactions costs, 209, 214
transnational (defined), 32
transnational advocacy networks, 14,
71, 86, 151, 152, 155, 157 8, 166,
171, 246
See also NGOs
transparency, 214
and social adaptation, 200
of society, 201
triangulation, 23 4, 61
tribal systems, 79
truth commissions, 25
Tumelty, Paul, 43
Tunisia, 231
Turkey, 43
and Chechnya, 60
and Cyprus, 173
and human rights, 81, 84 6
and Kurds, 87, 88, 175 6, 188, 192,
199 200, 219, 228, 256
and NATO, 88, 245
typological theory, 14, 216, 221
Ucarer, Emek, 176
Udugov, Movladi, 45
Uganda, xiii, 14, 228, 249 51, 148
and Sudan, 171, 217, 223
305
and Tutsis, 107 14, 118, 223,
226 7, 248
Gulu, 129, 132 3, 146
Operation Lightning Thunder, 146
Operation North, 129
Uighurs, 192 3
Umarov, Doku, 31, 45 6, 51 3, 55, 60 1
unions, 74, 83
United Nations, 86, 109, 112,
146, 167
and Afghanistan, 106
Boutros Boutros Ghali, 180
Department of Disarmament Affairs,
180
UNHCR, 90, 102, 111, 118
UNICEF, 156, 160, 170
UNOCHA, 146
United States
and Afghanistan, 97, 104, 106, 248
and al Qaeda, 231
and Chechnya, 31, 60
and Iraq, 222
and Irish diaspora, 71
and Kurds, 88
and LRA, 122, 125, 132, 146
and Nicaragua, 252, 256
and Sudan, 149, 151, 158, 160 6,
169 72, 225, 246 7, 254, 257
scholarship, 205
UPDF, 140 1, 145
Uzbekistan, 59
Uzbeks, 103
Vidino, Lorenzo, 47
violence, post return, 89, 92, 95, 99,
116, 188
wages, 83, 236
Wahabis, 31, 48 50, 52, 105
Walter, Barbara, 71
weapons. See under technology;
trafficking
Weidmann, Nils, 15, 123, 217, 219 20,
252 5
Wight, Colin, 205
Williams, Paul, 180
Wimmer, Andreas, 240
Wolf, Frank, 163 4
Wood, Elisabeth, 17, 19 20, 26, 92,
216, 220
306
Woodward, Peter, 170
World Bank, 116
Wucherpfennig, Julian, 240 1
Xinjiang, 192 3
Yandarbiyev, Zelimkhan, 44 5, 50 1
Yemen, 231
Index
Yousaf, Mohammad, 100
Zaire, 107 8, 113, 116, 119,
226 7
See also Congo, Democratic
Republic of
Zolberg, Aristide, 96
Zukerman, Zara, 34, 36